2.845

2023影响因子

(CJCR)

  • 中文核心
  • EI
  • 中国科技核心
  • Scopus
  • CSCD
  • 英国科学文摘

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于深度学习的视频超分辨率重建算法进展

唐麒 赵耀 刘美琴 姚超

王冰洁, 徐磊, 林宗利, 施阳, 杨涛. 基于自适应动态规划的量化通信下协同最优输出调节. 自动化学报, 2025, 51(4): 1−11 doi: 10.16383/j.aas.c240494
引用本文: 唐麒, 赵耀, 刘美琴, 姚超. 基于深度学习的视频超分辨率重建算法进展. 自动化学报, xxxx, xx(x): x−xx doi: 10.16383/j.aas.c240235
Wang Bing-Jie, Xu Lei, Lin Zong-Li, Shi Yang, Yang Tao. Cooperative optimal output regulation under quantized communication based on adaptive dynamic programming. Acta Automatica Sinica, 2025, 51(4): 1−11 doi: 10.16383/j.aas.c240494
Citation: Tang Qi, Zhao Yao, Liu Mei-Qin, Yao Chao. A review of video super-resolution algorithms based on deep learning. Acta Automatica Sinica, xxxx, xx(x): x−xx doi: 10.16383/j.aas.c240235

基于深度学习的视频超分辨率重建算法进展

doi: 10.16383/j.aas.c240235 cstr: 32138.14.j.aas.c240235
基金项目: 中央高校基本科研业务费专项资金资助(2024JBZX001), 国家自然科学基金(62120106009, 62332017, 62372036) 资助
详细信息
    作者简介:

    唐麒:北京交通大学信息科学研究所硕士研究生. 主要研究方向为图像与视频复原. E-mail: qitang@bjtu.edu.cn

    赵耀:北京交通大学信息科学研究所教授. 主要研究方向为图像/视频压缩, 数字媒体内容安全, 媒体内容分析与理解, 人工智能. E-mail: yzhao@bjtu.edu.cn

    刘美琴:北京交通大学信息科学研究所教授. 主要研究方向为多媒体信息处理, 三维视频处理, 视频智能编码. 本文通信作者. E-mail: mqliu@bjtu.edu.cn

    姚超:北京科技大学计算机与通信工程学院副教授. 主要研究方向为图像/视频压缩, 计算机视觉和人机交互. E-mail: yaochao@ustb.edu.cn

A Review of Video Super-resolution Algorithms Based on Deep Learning

Funds: Supported by Fundamental Research Funds for the Central Universities (2024JBZX001), and National Natural Science Foundation of China (62120106009, 62332017, 62372036)
More Information
    Author Bio:

    TANG Qi  Master student at Institute of Information Science, Beijing Jiaotong University. His research interest covers image and video restoration

    ZHAO Yao  Professor at Institute of Information Science, Beijing Jiaotong University. His research interest covers image/video compression, digital media content security, media content analysis and understanding, artificial intelligence

    LIU Mei-Qin  Professor at Institute of Information Science, Beijing Jiaotong University. Her research interest covers multimedia information processing, 3D video processing and video intelligent coding. Corresponding author of this paper

    YAO Chao  Associate Professor at School of Computer and Communication Engineering, University of Science and Technology Beijing. His research interest covers image/video compression, computer vision, and human–computer interaction

  • 摘要: 视频超分辨率重建(Video super-resolution, VSR)是底层计算机视觉任务中的一个重要研究方向, 旨在利用低分辨率视频的帧内和帧间信息, 重建具有更多细节和内容一致的高分辨率视频, 有助于提升下游任务性能和改善用户观感体验. 近年来, 基于深度学习的视频超分辨率重建算法如雨后春笋般涌现, 在帧间对齐、信息传播等方面取得了突破性的进展. 在简述视频超分辨率重建任务的基础上, 梳理了现有的视频超分辨率重建的公共数据集及相关算法; 接着, 重点综述了基于深度学习的视频超分辨率重建算法的创新性工作进展情况; 最后, 总结了视频超分辨率重建算法面临的挑战及未来的发展趋势.
  • 近年来, 多智能体系统的输出调节问题因其在无人机编队控制、自动驾驶和车联网以及多航天器姿态同步等领域的应用而引起广泛的关注[13]. 多智能体输出调节问题的目标是通过设计一种分布式控制策略, 实现每个跟随者的输出信号跟踪参考信号, 并抑制由外部系统描述的干扰信号[46]. 目前, 分布式控制策略的设计方法主要有两种: 前馈−反馈方法[78]与内模原理方法[910].

    此外, 在多智能体系统中, 智能体的通信通常受限于系统的通信拓扑结构, 智能体通常只能与邻居进行直接通信. 在领导−跟随多智能体系统中, 跟随者为获得领导者的状态信息, 可通过设计分布式观测器进行估计[7, 11]. 在自主水下航行器[12], 航天器编队控制[13]等实际网络通信中, 通信信道的有限带宽在智能体之间的信息传输中不容忽视[1418]. 为降低通信负担, 减少通信信道中传输数据的比特数, 一些学者通过设计量化器与编码−解码方案来解决量化通信下多智能体系统的协同输出调节问题. 文献[19]利用对数量化器对控制输入进行量化, 并通过扇形约束方法来处理存在的量化误差. 文献[20]通过设计一种基于缩放函数策略的动态编码−解码方案, 保证量化误差的收敛, 实现多智能体系统跟踪误差渐近收敛到零. 文献[21]将上述结果推广到具有切换拓扑图的多智能体系统上, 解决带有切换图的线性多智能体系统的量化协同输出调节问题. 值得注意的是, 上述研究中所设计的控制策略都是基于模型的, 这就要求每个智能体需要知道系统的模型信息. 然而, 由于通信带宽的固有限制和网络系统固有的脆弱性将导致如时间延迟, 数据包丢失, 信号量化以及网络攻击等现象的发生, 智能体难以完整获得整个系统的动态信息[2224].

    随着自适应动态规划的发展[2528], 一种针对不确定动态系统的自适应控制方法应运而生, 其优势在于可以利用在线数据通过学习来逼近动态系统的控制策略, 而不必完全了解系统的动态信息, 为模型未知的协同输出调节问题提供新的解决方案. 近年来, 一些学者将最优控制理论与自适应动态规划方法进行结合[2931], 通过数据驱动的方式求解最优/次优控制策略, 在保证闭环系统实现输出调节的同时, 最小化系统性能指标. 文献[3]利用前馈−反馈方法设计分布式控制策略, 解决跟随者对领导者状态未知的多智能体系统的协同最优输出调节问题. 文献[32]构建分布式自适应内部模型来估计领导者的动态, 并提出基于策略迭代与值迭代的强化学习算法, 在线学习最优控制策略. 文献[33]针对包含外部系统在内的所有智能体动态未知的多智能体系统, 利用内模原理与自适应动态规划方法, 解决协同最优输出调节问题. 然而, 上述的这些研究并未考虑通信信道带宽有限的情况. 而在实际的工程应用中, 如智能交通系统中的自适应巡航控制等问题, 往往期望设计一种能在通信带宽有限且系统动力学未知情况下运行的数据驱动算法, 来实现多智能体系统间的协同最优输出调节, 这促使我们对这一问题进行研究.

    本文的主要贡献如下: 1) 通过引入均匀量化器, 设计分布式量化观测器来减少通信信道中传输数据的比特数, 降低多智能体间的通信负担. 同时, 将均匀量化器引入到编码−解码方案设计中, 消除量化误差对多智能体系统的影响, 保证每个跟随者对外部系统状态的估计误差渐近收敛至零. 2) 将分布式量化观测器的估计值引入到次优控制策略的设计中, 在系统动态未知的情况下, 提出一种基于自适应动态规划的数据驱动算法, 在线学习次优控制策略, 解决量化通信下的协同最优输出调节问题. 3) 受文献[32]的启发, 在学习阶段, 本文考虑一个更一般的情况, 即跟随者系统只能通过观测器对领导者的状态进行估计, 而无法直接获得领导者的状态. 在这种情况下, 证明学习到的控制器增益将收敛到最优控制增益的任意小邻域内. 与现有文献相比, 文献[32]需要智能体间的精确通信, 而本文中智能体间传输的为量化后的信息, 降低了多智能体间的通信负担, 并通过引入编码−解码方案消除量化误差的影响, 实现量化通信下外部系统状态估计误差的渐近收敛. 文献[3, 34]不仅需要智能体间的精确通信, 并且需要假设每个跟随者系统都能够获得外部系统状态的实际值. 本文在学习阶段考虑一个更一般的情况, 跟随者系统可通过设计的分布式量化观测器对领导者的状态进行估计, 从而获得外部系统状态的估计值.

    本文其余部分安排如下. 第1节介绍图论的基础知识以及相关符号说明; 第2节介绍本文的问题描述; 第3节设计量化通信下的分布式观测器; 第4节提出自适应次优控制策略与自适应动态规划算法; 第5节在智能车联网自适应巡航控制系统上验证理论结果; 第6节总结本文的主要结果, 并提出未来的研究方向.

    本节介绍一些图论的基础知识以及相关符号的定义.

    多智能体系统通过通信网络与相邻的智能体之间共享信息, 该网络可以使用图论来描述. 在这一部分, 首先介绍图论的一些基本知识. 考虑一个具有$ N $个智能体的有向图$ \mathcal{G}=(\mathcal{V},\; \mathcal{E}) $, 其中$ \mathcal{V}= \{1,\;2,\;\cdots,\;N\} $表示智能体的集合, $ \mathcal{E} \subseteq \mathcal{V} \times \mathcal{V} $表示边的集合, 邻接矩阵被定义为$ \mathcal{A}=[a_{ij}] \in \bf{R}^{N\times N} $, 其中当$ a_{ij}> 0 $时, $ (j,\;i) \in \mathcal{E} $, 否则$ a_{ij}=0 $. 有向图$ \mathcal{G} $的拉普拉斯矩阵被定义为$ \mathcal{L}=[\ell_{ij}]\in \bf{R}^{N\times N} $, 其中$ \ell_{ii}=\sum\nolimits_{j=1}^{N}a_{ij} $, $ \ell_{ij}=-a_{ij} $, $ j\ne i $. 领导者由智能体$ 0 $表示, 由$ N $个智能体和领导者组成的图称为增广有向图$ \mathcal{\bar{G}}=(\mathcal{\bar{V}},\;\mathcal{\bar{E}}) $, 其中$ \mathcal{\bar{V}}= \{0,\;1,\;2,\;\cdots,\;N\} $表示智能体的集合, $ \mathcal{\bar{E}} \subseteq \mathcal{\bar{V}} \times \mathcal{\bar{V}} $表示边的集合. 如果从领导者智能体$ 0 $到智能体$ i\; \in\mathcal{V} $存在有向边, 则$ a_{i0}=1 $, 否则$ a_{i0}=0 $. 定义$ G={\rm diag}\{a_{10}, \;a_{20},\;\cdots,\; a_{N0}\} $表示对角矩阵, 令$ H=\mathcal{L}+G $, $ \mathcal{F}=H+\mathcal{A} $. $ \mathcal{N}_{i}=\left\{j|a_{ij}>0,\; j \in \mathcal{\bar{V}}\right\} $表示智能体 $ i\; \in\mathcal{V} $的邻居集合. 对于一个根节点而言, 如果存在从根节点到每个其他节点的有向路径, 则该有向图具有有向生成树.

    $ \bf{Z} $表示整数的集合. $ ||\cdot|| $为向量的欧氏范数和矩阵的$ 2 $范数. 对于列向量$ l=(l_{1},\; l_{2},\;\cdots,\; l_{n})^{{\mathrm{T}}} \in \bf{R}^{n} $, $ ||l||_{\infty}={\rm max}_{1\leq i\leq n}|l_{i}| $. $ \otimes $表示克罗内克积算子. 对于矩阵$ X \in \bf{R}^{m\times m} $, $ \rho(X) $表示它的谱半径, $ \lambda(X) $表示它的特征值, $ \sigma(X) $表示它的谱. $ {\rm tr}(X) $表示它的迹. $ X>0 $表示为正定矩阵, $ X\ge0 $表示为半正定矩阵. 对于矩阵$ X \in \bf{R}^{m\times n} $, $ {\rm rank}(X) $表示它的列秩. $ {\rm vec}(A)=[a^{{\mathrm{T}}}_{1},\; a^{{\mathrm{T}}}_{2},\; \cdots,\; a^{{\mathrm{T}}}_{q}]^{{\mathrm{T}}} \in \bf{R}^{pq} $ 表示将矩阵$ A\in \bf{R}^{p\times q} $向量化, 其中$ a_{i}\in\bf{R}^{p} $是矩阵$ A $的第$ i $列. 对于对称矩阵$ B \in \bf{R}^{m\times m} $, $ b_{mm} $为矩阵$ B $中第$ m $行第$ m $列的元素, $ {\rm vecs}(B)=[b_{11},\; 2b_{12},\;\cdots,\; 2b_{1m},\; b_{22}, 2b_{23},\;\cdots,\;2b_{m-1,\;m},\;b_{mm}]^{{\mathrm{T}}} \in \bf{R}^{\frac{1}{2}m(m+1)} $. 针对任意的列向量$ c\in \bf{R}^{n} $, $ c_{n} $为$ c $中第$ n $个元素, $ {\rm vecv}(c)= [c^{2}_{1},\;\, c_{1}c_{2},\;\,\cdots,\;\,c_{1}c_{n},\;\,c^{2}_{2},\;\,c_{2}c_{3},\;\cdots,\;c_{n-1}c_{n} $, $ c^{2}_{n}]^{{\mathrm{T}}} \in \bf{R}^{\frac{1}{2}n(n+1)}$. $ D={\rm blockdiag}\{D_{1},\;D_{2},\;\cdots,\;D_{N} \} $表示分块对角矩阵, 其中$ D_{i} $为对角块, $ i=1,\; 2,\;\cdots,\; N $. $ \mathbf{1}_n $与$ {I}_n $分别表示$ n $维全1列向量与$ n\times n $维单位矩阵. 针对复数$ {\textit z} $, $ {\rm Re}({\textit z}) $表示$ {\textit z} $的实部.

    本文考虑如下一类连续时间线性多智能体系统:

    $$ \dot{x}_i=A_{i}x_{i}+B_{i}u_{i}+D_{i}\omega\; $$ (1a)
    $$ \dot{\omega}=E\omega\; $$ (1b)
    $$ e_{i}=C_{i}x_{i}+F_{i}\omega,\; \quad i\in \mathcal{V}\; $$ (1c)

    其中, $ x_i\in\bf{R}^{n_i} $, $ u_i\in\bf{R}^{m_i} $, $ e_i\in\bf{R}^{p_i} $分别表示第$ i $个智能体的状态向量, 输入向量以及跟踪误差. 系统(1)的矩阵维数分别为$ A_i\in\bf{R}^{n_i\times n_i} $, $ B_i\in\bf{R}^{n_i\times m_i} $, $ D_i\in\bf{R}^{n_i\times q} $, $ C_i\in\bf{R}^{p_i\times n_i} $, $ F_i\in\bf{R}^{p_i\times q} $. 自治系统(1b)称为外部系统, 其中, $ \omega\in\bf{R}^{q} $表示外部系统的状态, $ E\in\bf{R}^{q\times q} $表示外部系统矩阵.

    针对以上系统, 本文给出一些基本假设条件如下所示:

    假设1. $ (A_i,\;B_i) $可镇定, $ i\in \mathcal{V} $.

    假设2. $ {\rm rank}\left[ \begin{matrix} A_{i}-\lambda I_{n_i} & B_{i} \\ C_{i} & 0 \end{matrix} \right]= n_{i}+p_{i},\; \forall \lambda \in \sigma(E),\; i\in \mathcal{V}. $

    假设3. 有向图$ \mathcal{\bar{G}} $包含以智能体$ 0 $为根节点的有向生成树.

    注1. 假设1和假设2均为多智能体系统输出调节问题中的基本假设[4, 30]. 如果假设3成立, 则$ H $的所有特征值均具有正实部[8].

    引理1[3, 8] . 假设1 ~ 3成立, 对于$ j=1,\;2,\;\cdots,\;q $, $ i\in \mathcal{V} $, 选择充分大的 $ \alpha>0 $ 使 $ {\rm Re}(\lambda_{j}(E)- \alpha\lambda_{i} (H))< 0 $, 其中$ \lambda_{j}(E) $和$ \lambda_{i}(H) $分别为$ E $的第$ j $个和$ H $的第$ i $个特征值, 令$ K_{i} $使$ A_{i}-B_{i}K_{i} $赫尔维玆, $ L_{i}=K_{i}X_{i}+U_{i} $, 其中$ (X_{i},\;U_{i}) $为以下调节器方程的一组解:

    $$ X_{i}E=A_{i}X_{i}+B_{i}U_{i}+D_{i}\; $$ (2a)
    $$ 0=C_{i}X_{i}+F_{i} $$ (2b)

    通过设计控制策略$ u_{i}=-K_{i}x_{i}+L_{i}\eta_{i} $可实现多智能体系统(1)的协同输出调节, 其中$ \eta_{i} $为第$ i $个跟随者对领导者状态$ \omega $的估计值.

    本文的控制目标是通过设计一种次优控制策略

    $$ u_{i}=-K^{*}_{i}x_{i}+L^{*}_{i}\eta_{i},\;\quad i\in \mathcal{V}\; $$ (3)

    实现多智能体系统的协同最优输出调节. 其中$ K^{*}_{i} $为最优反馈控制增益, $ L^{*}_{i} $为最优前馈控制增益.

    此外, 所设计的次优控制策略, 不仅需要解决协同输出调节问题, 同时还需要解决以下两个优化问题.

    问题1.

    $$ \begin{aligned} &\min\limits_{(X_{i},\;U_{i})}\quad {\rm tr}(X^{{\mathrm{T}}}_{i}Q_{i}X_{i}+U^{{\mathrm{T}}}_{i}R_{i}U_{i})\;\\ &\; \rm{s.t.}\quad (2)\; \end{aligned} $$

    其中, $ Q_{i}=Q^{{\mathrm{T}}}_{i}>0 $, $ R_{i}=R^{{\mathrm{T}}}_{i}>0 $.

    根据文献[35]可知, 求解静态优化问题1能够得到调节器方程(2)的唯一最优解$ (X^{*}_{i},\;U^{*}_{i}) $, 最优前馈控制增益$ L^{*}_{i}=K^{*}_{i}X^{*}_{i}+U^{*}_{i} $. 接下来, 为得到最优反馈控制增益$ K^{*}_{i} $, 需要求解以下动态规划问题.

    定义状态误差变量$ \bar{x}_{i}=x_{i}-X^{*}_{i}\omega $与输入误差变量$ \bar{u}_{i}=u_{i}-U_{i}^{*}\omega $. 根据调节器方程(2)与次优控制策略(3)能够得到系统(1a)的误差系统为

    $$ \dot{\bar{x}}_{i}=A_{i}\bar{x}_{i}+B_{i}\bar{u}_{i}\; $$ (4a)
    $$ e_{i}=C_{i}\bar{x}_{i}\; $$ (4b)

    其中, 控制输入为$ \bar{u}_{i}=-K^{*}_{i}\bar{x}_{i}+L^{*}_{i}(\eta_{i}-\omega) $. 误差系统(4)的最优控制策略为$ \bar{u}_{i}=-K^{*}_{i}\bar{x}_{i} $, 可通过求解以下优化问题获得.

    问题2.

    $$ \begin{aligned} &\min \limits_{\bar{u}_{i}}\quad \int_{0}^{\infty} (\bar{x}^{{\mathrm{T}}}_{i}\bar{Q}_{i}\bar{x}_{i}+\bar{u}^{{\mathrm{T}}}_{i}\bar{R}_{i}\bar{u}_{i}){\mathrm{d}}t\;\\ &\; \rm{s.t.}\quad (4)\; \end{aligned} $$

    其中, $ \bar{Q}_{i} = \bar{Q}^{{\mathrm{T}}}_{i}\ge 0 $, $ \bar{R}_{i} = \bar{R}^{{\mathrm{T}}}_{i}>0 $, $ (A_{i},\;\sqrt{\bar{Q}_{i}}) $可观测.

    问题2是一个标准的线性二次型调节器问题, 根据线性最优控制理论, 最优反馈增益$ K^{*}_{i} $为

    $$ K^{*}_{i}=\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}P^{*}_{i}\; $$ (5)

    其中, $ P^{*}_{i}=(P^{*}_{i})^{{\mathrm{T}}}>0 $是以下代数黎卡提方程的唯一解:

    $$ A^{{\mathrm{T}}}_{i}P_{i}^{*}+P_{i}^{*}A_{i}+\bar{Q}_{i}-P_{i}^{*}B_{i}\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}P_{i}^{*}=0 $$ (6)

    注2. 根据文献[3]中定理1的分析可知, 问题2的性能指标中应用控制策略$ \bar{u}_{i}=-K^{*}_{i}\bar{x}_{i}+L^{*}_{i}(\eta_{i}\,- \omega) $与最优控制策略$ \bar{u}_{i}=-K^{*}_{i}\bar{x}_{i} $之间的成本误差是有界的. 因此, 本文所设计的控制策略(3)是次优控制策略.

    由于最优反馈控制增益$ K^{*}_{i} $和最优前馈控制增益$ L^{*}_{i} $是相互独立的, 因此问题1和问题2可以分别进行求解. 值得注意的是, 直接求解非线性方程(6)往往比较困难, 尤其是针对维数比较高的矩阵. 因此, 通常采用以下策略迭代的方法来解决此类问题[36].

    简单而言, 选择一个使闭环系统稳定并保证所需成本有限的反馈控制增益$ K_{i,\;0} $, 即$ A_{i}-B_{i}K_{i,\;0} $是赫尔维玆矩阵. 通过策略迭代的方式求解如下Lyapunov方程来更新值$ P_{i,\;k} $:

    $$ \begin{split} &(A_{i}-B_{i}K_{i,\;k})^{{\mathrm{T}}}P_{i,\;k}+P_{i,\;k}(A_{i}-B_{i}K_{i,\;k})\;+\\ & \qquad\bar{Q}_{i}+ K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i}K_{i,\;k}=0\; \end{split} $$ (7)

    其中, $ k=1,\;2,\;\cdots $表示迭代次数. 通过以下方程来更新反馈控制增益

    $$ K_{i,\;k+1}=\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}P_{i,\;k} $$ (8)

    文献[36]已证明策略迭代方法中的每一次迭代反馈控制增益$ K_{i,\;k} $都可接受, 即保证了$ A_{i}\;- B_{i}K_{i,\;k} $是赫尔维玆矩阵. 同时也保证了$ \mathop{\lim}\nolimits_{k \to \infty}K_{i,\;k} = K_{i}^* $且$ \mathop{\lim}\nolimits_{k \to \infty}P_{i,\;k}=P_{i}^* $.

    为降低多智能体间的通信负担, 在本节中, 通过引入量化器与编码−解码方案设计分布式量化观测器, 用于估计量化通信下领导者的状态$ \omega $.

    在正式介绍编码−解码方案之前, 首先考虑一种均匀量化器$ \mathcal{Q}[e] $[37]:

    $$ \mathcal{Q}[e]=\varsigma,\;\quad \varsigma-\frac{1}{2}<e \leq \varsigma+\frac{1}{2}\; $$ (9)

    其中, $ \varsigma\in\bf{Z} $, $ e $表示需要量化的变量.

    给定向量$ h=[h_{1},\;h_{2}\cdots,\;h_{n}]\in \bf{R}^{n} $, 定义量化器$ \mathcal{Q}[h]=[\mathcal{Q}[h_{1}],\;\cdots,\; \mathcal{Q}[h_{n}]] $. 量化误差为

    $$ ||h-\mathcal{Q}[h]||_{\infty} \leq \frac{1}{2} $$ (10)

    由于量化误差的存在, 智能体无法获得邻居传输的准确信息, 为消除量化误差带来的影响, 将量化器引入到如下编码−解码方案的设计之中.

    1)编码器

    为传输$ \eta_j(k) $量化后的数据, 对于任意$ k\ge1 $, 智能体$ j \in \mathcal{\bar{V}} $生成的量化输出为$ {\textit z}_j(k) $, 即

    $$ {\textit z}_{j}(k)=\mathcal{Q}\left[\frac{1}{s(k-1)}(\eta_j(k)-b_j(k-1))\right]\; $$ (11a)
    $$ b_j(k)=s(k-1){\textit z}_{j}(k)+b_j(k-1) $$ (11b)

    其中, 内部状态$ b_j(k) $的初值$ b_j(0)=0 $, $ s(k)= s(0) \mu^k>0 $为自适应调整编码器的递减序列, $ \mu\in (0,\;1) $.

    2)解码器

    当智能体$ i $从邻居智能体$ j $接收到量化后的数据$ {\textit z}_{j}(k) $时, 通过以下规则递归生成$ \eta_j(k) $的估计值$ \hat{\eta}_j(k) $, 并通过零阶保持器输出为连续信号$ \hat{\eta}_j(t) $, 即

    $$ \hat{\eta}_j(k)=s(k-1){\textit z}_{j}(k)+\hat{\eta}_j(k-1)\; $$ (12a)
    $$ \hat{\eta}_j(t)=\hat{\eta}_j(k),\; kT \leq t<(k+1)T\; $$ (12b)

    其中, 初值$ \hat{\eta}_j(0)=0 $, $ T>0 $为采样时间, 其选取遵循香农采样定理.

    图 1所示, 对智能体$ i $和邻居智能体$ j $之间的通信而言, 在每个采样时刻, 智能体$ j $对外部系统状态的估计值$ \eta_j(t) $进行采样, 并将采样后的数据$ \eta_j(k) $编码为量化后的数据$ {\textit z}_j(k) $, 然后通过通信信道传输给邻居智能体$ i $. 邻居智能体$ i $接收到数据信息之后通过解码器解码为$ \hat{\eta}_j(k) $, 进而通过零阶保持器得到发送者状态的估计值$ \hat{\eta}_j(t) $. 其中$ b_j(k) $表示一个预测器, 目的是预测智能体$ j $的邻居智能体$ i $经过解码后的得到的数据$ \hat{\eta}_j(k) $.

    图 1  编码−解码方案
    Fig. 1  Encoder-decoder scheme

    注3. 在编码−解码方案设计中, $ s(k) $表示用于调整预测误差$ \eta_j(k)-b_j(k-1) $的调节函数. $ \mu\in (0,\;1) $保证了随着迭代次数的增加, 智能体$ i $对邻居智能体$ j $传输数据的估计误差$ \eta_j(k)-\hat{\eta}_j(k) $逐渐减小, 即消除了量化误差对传输数据准确性的影响.

    接下来, 将上述经编码−解码方案传输的估计值$ \hat{\eta}_j(t) $引入到分布式观测器的设计当中, 针对每个跟随者$ i \in \mathcal{V} $, 受文献[8]的启发, 本文构建分布式量化观测器如下:

    $$ \dot{\eta}_i=E\eta_i+\alpha \sum\limits_{j \in \mathcal{N}_i} a_{i j}\left(\hat{\eta}_j-\eta_i\right) $$ (13)

    其中, $ \eta_i \in \bf{R}^{q} $, 参数$ \alpha>0 $. $ \hat{\eta}_j \in \bf{R}^{q} $表示智能体$ i $对$ \eta_j $经过编码−解码后的估计值, $ \hat{\eta}_0 = \hat{\omega} $.

    本文理论部分的全文流程图如图 2所示. 本文利用量化器与编码−解码方案设计分布式量化观测器, 在减少通讯负担的同时, 对外部系统的状态进行估计. 定理1证明了所提观测器对外部系统状态估计误差的收敛性. 通过求解问题1与问题2设计次优控制策略. 当系统模型未知时, 我们给出一个在线学习算法1, 通过数据驱动的方式在线求解次优控制策略. 定理2则证明了由算法1得到的次优控制策略能够实现量化通信下的自适应协同最优输出调节.

    图 2  理论部分示意图
    Fig. 2  Illustration of the theoretical part

    接下来, 通过以下定理说明所设计的分布式量化观测器保证了对外部系统状态估计误差的收敛性.

    定理1. 考虑外部系统(1b)和分布式量化观测器(13), 如果假设1 ~ 3成立, 对于充分大的$ \alpha>0 $, 经过编码−解码后, 智能体$ i $对外部系统状态的估计误差

    $$ \mathop{\lim}\limits_{t \to \infty}(\eta_{i}(t)-\omega(t))=0\; $$ (14)

    其中, $ i \in \mathcal{V} $.

    证明. 定义$ \bar{\eta}(t)=[\eta_{1}^{{\mathrm{T}}}(t),\; \eta_{2}^{{\mathrm{T}}}(t),\; \cdots,\; \eta_{N}^{{\mathrm{T}}}(t)]^{{\mathrm{T}}} $, $ \hat{\eta}(t)=[\hat{\eta}_{1}^{{\mathrm{T}}}(t),\; \hat{\eta}_{2}^{{\mathrm{T}}}(t),\; \cdots,\; \hat{\eta}_{N}^{{\mathrm{T}}}(t)]^{{\mathrm{T}}} $, $ \bar{\omega}(t)=\mathbf{1}_N\otimes \omega(t) $, $ \hat{\bar{\omega}}(t)=\mathbf{1}_N\otimes\hat{\omega}(t) $, $ \bar{E}={ I_{{N}}}\otimes E $. 将外部系统(1b)与分布式量化观测器(13)整理成如下紧凑形式:

    $$ \dot{\bar{\omega}}(t)=\bar{E}\bar{\omega}(t)\; $$ (15a)
    $$ \begin{split} \dot{\bar{\eta}}(t)=\;&\bar{E}\bar{\eta}(t)-\alpha(\mathcal{F}\otimes I_{q})\bar{\eta}(t)\;+ \\ &\alpha(\mathcal{A}\otimes I_{q})\hat{\eta}(t)+\alpha(H\otimes I_{q})\hat{\bar{\omega}}(t) \end{split} $$ (15b)

    定义$ e_{\omega}(t)=\bar{\omega}(t)-\hat{\bar{\omega}}(t) $, $ e_{\eta}(t)=\bar{\eta}(t)-\hat{\eta}(t) $, 可得

    $$ \begin{split} \dot{\bar{\eta}}(t)=\;&(\bar{E}-\alpha(H\otimes I_{q}))\bar{\eta}(t)\;+\\ &\alpha(H\otimes I_{q})\bar{\omega}(t)-\alpha(\mathcal{A}\otimes I_{q})e_{\eta}(t)\;-\\ &\alpha(H\otimes I_{q})e_{\omega}(t) \end{split} $$ (16)

    定义$ \tilde{\eta}(t)=\bar{\eta}(t)-\bar{\omega}(t) $, 根据式(15a)和式(16)有

    $$ \begin{split} \dot{\tilde{\eta}}(t)=\;&(\bar{E}-\alpha(H\otimes I_{q}))\tilde{\eta}(t)\;-\\ &\alpha(\mathcal{A}\otimes I_{q})e_{\eta}(t)-\alpha(H\otimes I_{q})e_{\omega}(t) \end{split} $$ (17)

    根据引理1可知, 对于$ j=1,\;2,\;\cdots,\;q $, $ i\in \mathcal{V} $, $ {\rm Re}(\lambda_{j}(E)-\alpha\lambda_{i}(H))<0 $, 其中$ \lambda_{j}(E) $和$ \lambda_{i}(H) $分别为$ E $的第$ j $个和$ H $的第$ i $个特征值, 即$ \bar{E}- \alpha(H\otimes I_{q}) $是赫尔维玆的.

    令$ E_h=\bar{E}-\alpha(H\otimes I_{q}) $, $ E_H=\alpha(H\otimes I_{q}) $, $ E_A= \alpha(\mathcal{A}\otimes I_{q}) $, 则式(16)可改写为

    $$ \begin{split} \dot{\bar{\eta}}(t)=\;&E_{h}\bar{\eta}(t)+E_{H}\bar{\omega}(t)\;-\\ &E_{A}e_{\eta}(t)-E_{H}e_{\omega}(t) \end{split} $$ (18)

    由于$ \hat{\bar{\omega}}(t) $与$ \hat{\eta}(t) $使用编码−解码方案进行更新, 将系统(15a)与(18)进行离散化. 定义$ e_{\omega}(k)= \bar{\omega}(k)-\hat{\bar{\omega}}(k) $, $ e_{\eta}(k)=\bar{\eta}(k)-\hat{\eta}(k) $, 系统(15a)与(18)利用零阶保持器方法进行离散化[38], 即

    $$ \bar{\omega}(k+1)={\mathrm{e}}^{\bar{E}{{T}}}\bar{\omega}(k)\; $$ (19a)
    $$ \begin{split} \bar{\eta}(k+1)=\;&{\mathrm{e}}^{E_{h}{{T}}}\bar{\eta}(k)+\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{H}{\mathrm{d}}\tau\bar{\omega}(k)\; -\\ &\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{A}{\mathrm{d}}\tau e_{\eta}(k) \;-\\ &\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{H}{\mathrm{d}}\tau e_{\omega}(k)\; \end{split} $$ (19b)

    其中, $ T $为采样时间, 其选取遵循香农采样定理.

    接下来, 将预测器$ b_{j}(k) $表示为紧凑型, 其中$ j \in \mathcal{\bar{V}} $. 定义$ b_{\omega}(k)=\mathbf{1}_N\otimes b_0(k) $, $ b_{\eta}(k)=[b_1^{{\mathrm{T}}}(k),\;b_2^{{\mathrm{T}}} (k),\; \cdots,\; b_N^{{\mathrm{T}}}(k)]^{{\mathrm{T}}} $. 预测器$ b_{j}(k) $表示对智能体 $ i $经过解码后得到的数据$ \hat{\eta}_j(k) $的预测, 根据$ \hat{\eta}_0(k) = \hat{\omega}(k) $, 且初始值$ b_{\omega}(0)=\hat{\bar{\omega}}(0) $, $ b_{\eta}(0)=\hat{\eta}(0) $, 可得$ b_{\omega}(k)=\hat{\bar{\omega}}(k) $, $ b_{\eta}(k)=\hat{\eta}(k) $. 因此, $ e_{\omega}(k)= \bar{\omega}(k)\;- b_{\omega}(k) $, $ e_{\eta}(k)=\bar{\eta}(k)-b_{\eta}(k) $.

    根据式(11), 有

    $$ \begin{split} b_{\omega}(k)=\;&s(k - 1)\mathcal{Q}\left[\frac{1}{s(k - 1)}(\bar{\omega}(k) - b_{\omega}(k - 1))\right] +\\&b_{\omega}(k-1) \end{split} $$ (20a)
    $$ \begin{split} b_{\eta}(k)=\;&s(k - 1)\mathcal{Q}\left[\frac{1}{s(k - 1)}(\bar{\eta}(k) - b_{\eta}(k - 1))\right]+\\ &b_{\eta}(k-1) \end{split} $$ (20b)

    将式(19a)的左右两边同时减去$ b_{\omega}(k) $, 可以得到

    $$ \begin{split} &\bar{\omega}(k+1)-b_{\omega}(k)={\mathrm{e}}^{\bar{E}T}\bar{\omega}(k)-b_{\omega}(k)=\\ &\quad {{e}}_{\omega}(k)+({\mathrm{e}}^{\bar{E}T}-I_{qN})\bar{\omega}(k)=s(k)\theta_{\omega}(k)\; \end{split} $$ (21)

    其中, $ \theta_{\omega}(k)=\frac{e_{\omega}(k)}{s(k)}+\frac{1}{s(k)}({\mathrm{e}}^{\bar{E}T}-I_{qN})\bar{\omega}(k) $.

    基于式(20a)和式(21), 可得

    $$ \begin{split} e_{\omega}(k+1)=\;&\bar{\omega}(k+1)-b_{\omega}(k+1)= \\ & \bar{\omega}(k+1)-b_{\omega}(k)\;-\\ & s(k)\mathcal{Q}\left[\frac{1}{s(k)}(\bar{\omega}(k+1)-b_{\omega}(k))\right]=\\ & s(k)(\theta_{\omega}(k)-\mathcal{Q}[\theta_{\omega}(k)])\\[-3pt]\end{split} $$ (22)

    同理, 将式(19b)的左右两边同时减去$ b_{\eta}(k) $, 可得

    $$ \begin{split} &\bar{\eta}(k+1)-b_{\eta}(k)=\\ &\quad ({\mathrm{e}}^{E_{h}T}-I_{qN})\bar{\eta}(k)+\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{H}{\mathrm{d}}\tau\bar{\omega}(k)\;+\\ &\quad (I_{qN}-\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{A}{\mathrm{d}}\tau)e_{\eta}(k)\;-\\ &\quad \int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{H}{\mathrm{d}}\tau e_{\omega}(k)= s(k)\theta_{\eta}(k)\; \end{split} $$ (23)

    其中,

    $$\begin{split} \theta_{\eta}(k)=&\frac{1}{s(k)}({\mathrm{e}}^{E_{h}T}-I_{qN})\bar{\eta}(k)\;+\\&\frac{1}{s(k)}\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau} E_{H} {\mathrm{d}}\tau\bar{\omega}(k)\;+\\& \frac{e_{\eta}(k)}{s(k)}(I_{qN}-\int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau}E_{A}{\mathrm{d}}\tau)\;-\\& \frac{e_{\omega}(k)}{s(k)} \int_{0}^{{{T}}}{\mathrm{e}}^{E_{h}\tau} E_{H}{\mathrm{d}}\tau \end{split}$$

    基于式(20b)和式(23), 可得

    $$ \begin{split} e_{\eta}(k+1)=\;&\bar{\eta}(k+1)-b_{\eta}(k+1)=\\ & \bar{\eta}(k+1)-b_{\eta}(k)\;-\\ & s(k)\mathcal{Q}\left[\frac{1}{s(k)}(\bar{\eta}(k+1)-b_{\eta}(k))\right]=\\ & s(k)(\theta_{\eta}(k)-\mathcal{Q}[\theta_{\eta}(k)]) \end{split} $$ (24)

    根据式(22), 式(24)以及量化误差(10), 有

    $$ ||\frac{e_{\omega}(k)}{s(k)}||_{\infty}\leq\frac{1}{2\mu}\; $$ (25a)
    $$ ||\frac{e_{\eta}(k)}{s(k)}||_{\infty}\leq\frac{1}{2\mu}\; $$ (25b)

    由$ \mathop{\lim}\nolimits_{k \to \infty}s(k) = 0 $可知$ \mathop{\lim}\nolimits_{k \to \infty}e_{\omega}(k) = e_{\eta}(k) = 0 $, 进而可知$ \mathop{\lim}\nolimits_{t \to \infty}e_{\omega}(t) = e_{\eta}(t) = 0 $. 由$ \bar{E}-\alpha(H\otimes I_{q}) $是赫尔维玆的, $ \mathop{\lim}\nolimits_{t \to \infty}e_{\omega}(t)=e_{\eta}(t)=0 $, 根据文献[39]引理$ 9.1 $, 可知$ \mathop{\lim}\nolimits_{t \to \infty}\tilde{\eta}(t)=0 $. 因此, 对于每个跟随者$ i \in \mathcal{V} $, 有$ \mathop{\lim}\nolimits_{t \to \infty}\tilde{\eta}_{i}(t)=0 $.

    在第3节中, 通过设计的分布式量化观测器可使每个跟随者渐近观测到外部系统的状态信息. 在本节中, 将观测到的估计值$ \eta_{i}(t) $引入到自适应动态规划算法的学习阶段, 进而设计一种数据驱动的方法来解决量化通信下的协同最优输出调节问题. 值得注意的是, 该方法能够近似逼近控制增益$ K^* $与$ L^* $, 而不需要知道系统矩阵$ A_{i} $, $ B_{i} $与$ D_{i} $的先验知识.

    考虑第$ i $个智能体, 定义$ \bar{x}_{ij}=x_{i}-X_{ij}\omega $, $ X_{ij}\in \bf{R}^{n_{i}\times q} $表示$ C_{i}X_{ij}+F=0 $的基础解系. 其中, $ i \in \mathcal{V} $, $ j=0,\;1,\;\cdots,\;h_{i}+1 $. $ h_{i}=(n_{i}-p_{i })q $ 表示 $ I_{q}\otimes C_{i} $零空间的维数. 接下来, 定义一个西尔维斯特方程$ S_{i}(X_{ij})=X_{ij}E-A_{i}X_{ij} $, $ X_{ij} \in \bf{R}^{n_{i} \times q} $, 根据输入误差变量$ \bar{u}_{i}=u_{i}-U_{i}^{*}\omega $与(2a), 式(4)可改写为

    $$ \begin{split} \dot{\bar{x}}_{i}=&\;A_{i}\bar{x}_{i}+B_{i}\bar{u}_{i}=\\ &\bar{A}_{i}\bar{x}_{ij}+B_{i}(K_{i,\;k}\bar{x}_{ij}+u_{i})\;+\\ &(D_{i}-S_{i}(X_{ij}))\omega =\\ &\bar{A}_{i}\bar{x}_{ij}+B_{i}(K_{i,\;k}\bar{x}_{ij}+u_{i})\;+\\ & (D_{i}-S_{i}(X_{ij}))\eta_{i}-(D_{i}-S_{i}(X_{ij}))\tilde{\eta}_{i} \end{split} $$ (26)

    其中, $ \bar{A}_{i}=A_{i}-B_{i}K^{*}_{i} $. 通过增大$ \alpha $, 可使$ \tilde{\eta}_{i}(t) $以所需的速度收敛到零[32].

    根据式(26)以及代数黎卡提方程(7)和(8), 有

    $$ \begin{split} &\bar{x}^{{\mathrm{T}}}_{ij}(t+\delta)P_{i,\;k}\bar{x}_{ij}(t+\delta)-\bar{x}^{{\mathrm{T}}}_{ij}(t)P_{i,\;k}\bar{x}_{ij}(t)=\\ &\quad\int_{t}^{t+\delta} (\bar{x}^{{\mathrm{T}}}_{ij}(\bar{A}_{i}^{{\mathrm{T}}}P_{i,\;k}+P_{i,\;k}\bar{A}_{i})\bar{x}_{ij}\;+\\ &\quad2(u_{i}+K_{i,\;k}\bar{x}_{ij})^{{\mathrm{T}}}B^{{\mathrm{T}}}_{i}P_{i,\;k}\bar{x}_{ij}\;+\\ &\quad2\eta_{i}^{{\mathrm{T}}}(D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}P_{i,\;k}\bar{x}_{ij})\,\; {\mathrm{d}}\tau=\\ &\quad\int_{t}^{t+\delta} (-\bar{x}^{{\mathrm{T}}}_{ij}(\bar{Q}_{i}+ K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i}K_{i,\;k})\bar{x}_{ij}\;+\\ &\quad2(u_{i}+K_{i,\;k}\bar{x}_{ij})^{{\mathrm{T}}}\bar{R}_{i}K_{i,\;k+1}\bar{x}_{ij}\;+\\ &\quad2\eta_{i}^{{\mathrm{T}}}(D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}P_{i,\;k}\bar{x}_{ij})\,\; {\mathrm{d}}\tau \end{split} $$ (27)

    通过克罗内克积的性质, 有

    $$ \begin{split} &\bar{x}^{{\mathrm{T}}}_{ij}(\bar{Q}_{i}+ K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i}K_{i,\;k})\bar{x}_{ij}= \\ &\quad(\bar{x}^{{\mathrm{T}}}_{ij}\otimes \bar{x}^{{\mathrm{T}}}_{ij}){\rm vec}(\bar{Q}_{i}+ K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i}K_{i,\;k})\; \end{split} $$ (28a)
    $$ \begin{split} &(u_{i}+K_{i,\;k}\bar{x}_{ij})^{{\mathrm{T}}}\bar{R}_{i}K_{i,\;k+1}\bar{x}_{ij} =\\ &\quad((\bar{x}^{{\mathrm{T}}}_{ij}\otimes \bar{x}^{{\mathrm{T}}}_{ij})(I_{ni}\otimes K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i})\;+ \\ &\quad(\bar{x}^{{\mathrm{T}}}_{ij}\otimes u^{{\mathrm{T}}}_{i})(I_{ni}\otimes \bar{R}_{i})){\rm vec}(K_{i,\;k+1})\; \end{split} $$ (28b)
    $$ \begin{split} &\eta_{i}^{{\mathrm{T}}}(D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}P_{i,\;k}\bar{x}_{ij}= \\ &\quad(\bar{x}^{{\mathrm{T}}}_{ij}\otimes \eta_{i}^{{\mathrm{T}}}){\rm vec}((D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}P_{i,\;k}) \end{split} $$ (28c)

    对于任意两个向量$ p $, $ q $以及正整数$ c $, 定义以下矩阵

    $$ \begin{split} {\Pi}_{pp}=\;&[\mathrm{vecv}(p(t_{1}))-\mathrm{vecv}(p(t_{0})),\;\cdots,\; \\ & \mathrm{vecv}(p(t_{c}))-\mathrm{vecv}(p(t_{c-1}))]^{{\mathrm{T}}}\; \end{split} $$ (29a)
    $$ {\Xi}_{pq}=\left[\int_{t_{0}}^{t_{1}}p\otimes q {\mathrm{d}}\tau,\;\cdots,\;\int_{t_{c-1}}^{t_{c}}p\otimes q {\mathrm{d}}\tau \right]^{{\mathrm{T}}}\; $$ (29b)

    其中, $ t_0<t_1<\cdots<t_c $, 基于以上矩阵定义, 通过式(27)得到以下矩阵方程

    $$ \Psi_{ij,\;k} \begin{bmatrix} {\rm vecs}(P_{i,\;k}) \\ {\rm vec}(K_{i,\;k+1})\\ {\rm vec}((D_{i}-S_{i}(X_{ij})^{{\mathrm{T}}}P_{i,\;k}) \end{bmatrix} =\Phi_{ij,\;k} $$ (30)

    其中,

    $$ \begin{split} \Psi_{ij,\;k}=\;&[ \Pi_{\bar{x}_{ij}\bar{x}_{ij}},\; -2\Xi_{\bar{x}_{ij}\bar{x}_{ij}}(I_{ni}\otimes K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i}) \;-\\ & 2\Xi_{\bar{x}_{ij}u_{i}}(I_{ni}\otimes \bar{R}_{i}),\;-2\Xi_{\bar{x}_{ij}\eta_{i}}]\; \end{split} $$ (31a)
    $$ \Phi_{ij,\;k}= -\Xi_{\bar{x}_{ij}\bar{x}_{ij}} {\rm vec}(\bar{Q}_{i}+K^{{\mathrm{T}}}_{i,\;k}\bar{R}_{i}K_{i,\;k}) $$ (31b)

    如果矩阵$ \Psi_{ij,\;k} $列满秩, 则式(30)具有唯一解. 文献[30]引理$ 3 $中给出矩阵$ \Psi_{ij,\;k} $列满秩的充分条件. 如果存在正整数$ c^{*} $使得任意的$ c>c^{*} $和时间序列$ t_{0}<t_{1}<\cdots<t_{c} $, 满足以下条件时,

    $$ \begin{split}& {\rm rank}([\Xi_{\bar{x}_{ij}\bar{x}_{ij}},\;\Xi_{\bar{x}_{ij}u_{i}},\;\Xi_{\bar{x}_{ij}\eta_{i}}])=\\&\quad \frac{n_{i}(n_{i}+1)}{2}+(m_{i}+q)n_{i}\; \end{split} $$ (32)

    矩阵$ \Psi_{ij,\;k} $对任意正整数$ k $列满秩.

    根据调节器方程(2), 西尔维斯特方程$ S_{i}(X_{ij})= X_{ij}E-A_{i}X_{ij} $以及式(30)的解, 能够求得调节器方程的解$ (X_{i},\;U_{i}) $. 该方法与文献[3]中式(27)的求解思路一致, 这里不做赘述.

    为确保满秩条件(32)能够得到满足, 在学习阶段$ [t_{0},\;t_{c}] $, 本文在初始控制策略上增加探测噪声$ \xi_{i} $, 即$ u_{i0}=-K_{i0}x_{i}+\xi_{i} $, 其中, $ K_{i0} $使$ A_{i}-B_{i}K_{i0} $赫尔维玆.

    据此, 针对量化通信下的自适应协同最优输出调节问题, 本文给出一个在线学习算法, 即算法1.

    算法1. 基于自适应动态规划的量化通信下协同最优输出调节算法

    1: 令$ i=1 $

    2: 选择一个初始控制策略$ u_{i0}=-K_{i0}x_{i}+\xi_{i} $

    3: 通过式(13)计算编码−解码后对外部系统状态的估 计值$ \eta_{i} $

    4: 计算满足条件(32)的$ \Xi_{\bar{x}_{ij}\bar{x}_{ij}},\;\Xi_{\bar{x}_{ij}u_{i}},\;\Xi_{\bar{x}_{ij}\eta_{i}} $

    5: 令$ k=0 $

    6: 通过式(30)求解$ P_{i,\;k} $, $ K_{i,\;k+1} $以及$ S_{i}(X_{ij}) $

    7: 令$ k\gets k+1 $, 重复步骤6, 直至满足$ ||P_{i,\;k}\;- \qquad\qquad P_{i,\;k-1}||<c_{i} $, 其中, 阈值$ c_{i} $为足够小的正数

    8: $ k^{*}\gets k $

    9: $ P_{i,\;k^*}\gets P_{i,\;k} $, $ K_{i,\;k^*}\gets K_{i,\;k} $

    10: 通过$ S_{i}(X_{ij}) $以及问题1求解调节器方程的最优解    $ (X^{*}_{i},\;U^{*}_{i}) $, $ L_{i,\;k^*}=K_{i,\;k^*}X^{*}_{i}+U^{*}_{i} $

    11: 学习到的次优控制策略为

    $$ u_{i}^*=-K_{i,\;k^*}x_{i}+L_{i,\;k^*}\eta_{i}\; $$ (33)

    12: 令$ i\gets i+1 $, 重复步骤2 ~ 11, 直至$ i=N $.

    注4. 本文利用所设计的算法1通过系统状态$ x_{i} $, 输入$ u_{i} $以及对外部系统状态的估计值$ \eta_{i} $在线学习次优控制策略(3), 而不需要依赖系统矩阵$ A_{i} $, $ B_{i} $与$ D_{i} $的先验知识. 然而, 由于在分布式量化观测器的设计部分应用外部系统的矩阵信息, 因此要求跟随者对外部系统矩阵$ E $是已知的. 目前, 在精确通信下, 文献[7, 11]不要求跟随者对外部系统矩阵$ E $是已知的, 即已经研究了部分/全部跟随者无法访问领导者系统矩阵信息的情况, 并设计了自适应分布式观测器. 然而在量化通信下, 文献[7, 11]中所设计的自适应分布式观测器并不适用, 需要设计自适应分布式量化观测器对外部系统矩阵$ E $的估计值$ E_{i}(t) $进行观测, 其中观测器中包含经过编码−解码方案后传输的信息$ \hat{E}_{i}(t) $, 我们难以保证估计误差$ {\lim}_{t \to \infty}(E_{i}(t)-E) $收敛到零, 这对我们的研究带来全新的挑战, 在未来的工作中将进一步研究.

    接下来, 给出关于控制增益$ K_{i,\;k^*} $和值$ P_{i,\;k^*} $的收敛性的定理.

    定理2. 在满足条件(32)的情况下, 对于任意小的参数$ \delta>0 $, 存在充分大的$ \alpha>0 $使由算法1得到的解$ \left\{P_{i,\;k}\right\}_{k=0}^{\infty} $和$ \left\{K_{i,\;k}\right\}_{k=0}^{\infty} $满足不等式$ ||P_{i,\;k^*}- P_{i}^*||<\delta $, $ ||K_{i,\;k^*}-K_{i}^*||<\delta $, 其中$ i \in \mathcal{V} $. 且由算法1得到的次优控制策略能够实现量化通信下的协同最优输出调节.

    证明. 令$ \left\{\bar{P}_{i,\;k}\right\}_{k=0}^{\infty} $, $ \left\{\bar{K}_{i,\;k}\right\}_{k=0}^{\infty} $为基于模型迭代方法得到的解.

    基于模型方法的收敛性分析已经在文献[36]中得到证明. 对于每个跟随者$ i \in \mathcal{V} $, 存在$ k^* $使得以下不等式成立, 即

    $$ \begin{split}& ||\bar{K}_{i,\;k^*}-K_{i}^*||<\frac{\delta}{2}\;\\& ||\bar{P}_{i,\;k^*}-P_{i}^*||<\frac{\delta}{2} \end{split} $$ (34)

    接下来, 需要证明算法1在每次迭代中学到的控制增益$ K_{i,\;k} $和值$ P_{i,\;k} $足够接近基于模型算法(7)和(8)得到的控制增益$ \bar{K}_{i,\;k} $和值$ \bar{P}_{i,\;k} $. 下面将通过归纳法证明.

    当$ k=0 $时, 对于所有的跟随者$ i \in \mathcal{V} $, 有$ K_{i0}= \bar{K}_{i0} $. 定义$ \Delta P_{i0}=P_{i0}-\bar{P}_{i0} $. $ \Delta P_{i0} $可通过以下方程进行求解, 即

    $$ \begin{split}& \Psi_{ij,\;0} \begin{bmatrix} {\rm vecs}(\Delta P_{i0}) \\ {\rm vec}(\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}\Delta P_{i0})\\ {\rm vec}((D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}\Delta P_{i0})\\ \end{bmatrix}=\\&\qquad 2\Xi_{\bar{x}_{ij}\tilde{\eta}_{i}}{\rm vec}((D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}\bar{P}_{i0}) \end{split} $$ (35)

    令$ ||\Delta\tilde{\eta}||=\max\nolimits_{t_{0}\leq t\leq t_{c}}\tilde{\eta}(t) $, 可知

    $$\begin{split}& \lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0} (P_{i0}- \bar{P}_{i0})=0\\ &\lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0}(K_{i1}-\bar{K}_{i1})=\\&\qquad\lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0} (\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}(P_{i0}- \bar{P}_{i0}))=0 \end{split}$$

    当$ k=p $时, 假设$ \lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0}(K_{ip}-\bar{K}_{ip})=0 $. 令$ \Delta P_{ip}= P_{ip}-\bar{P}_{ip} $. $ \Delta P_{ip} $可通过以下方程进行求解

    $$ \Psi_{ij,\;0} \begin{bmatrix} {\rm vecs}(\Delta P_{ip}) \\ {\rm vec}(\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}\Delta P_{ip})\\ {\rm vec}((D_{i}-S_{i}(X_{ij}))^{{\mathrm{T}}}\Delta P_{ip}) \end{bmatrix} =\Delta \Phi_{ij,\;p} $$ (36)

    其中, $ \lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0}\Delta \Phi_{ij,\;p}=0 $. 因此, 可得

    $$\begin{split}&\lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0} (P_{ip}-\bar{P}_{ip})=0\\ &\lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0}(K_{i,\;p+1}- \bar{K}_{i,\;p+1})=\\& \qquad\lim\nolimits_{||\Delta\tilde{\eta}||\rightarrow0} (\bar{R}^{-1}_{i}B^{{\mathrm{T}}}_{i}(P_{ip}- \bar{P}_{ip}))=0 \end{split}$$

    通过增大$ \alpha $的值能够加速$ \Delta\tilde{\eta} $的收敛, 对于充分大的$ \alpha>0 $, 总能找到足够小的$ \Delta\tilde{\eta} $使得在任何迭代$ k $处, 满足不等式$ ||P_{i,\;k}-\bar{P}_{i,\;k}||<\delta/2 $, $ ||K_{i,\;k}\;- \bar{K}_{i,\;k}||<\delta/2 $.

    因此, 当$ k=k^* $时, 以下不等式成立, 即

    $$ \begin{split}& ||K_{i,\;k^*}-\bar{K}_{i,\;k^*}||<\frac{\delta}{2}\;\\& ||P_{i,\;k^*}-\bar{P}_{i,\;k^*}||<\frac{\delta}{2} \end{split} $$ (37)

    根据式(34)与式(37), 通过矩阵三角不等式可知, $ ||P_{i,\;k^*}-P_{i}^*||<\delta $, $ ||K_{i,\;k^*}-K_{i}^*||<\delta $.

    接下来, 证明由算法1得到的次优控制策略能够实现量化通信下的协同最优输出调节. 令$ \tilde{\eta}_{i}(t)= \eta_{i}(t)-\omega(t) $, 由定理1可知, 在量化通信, 对外部系统状态的估计误差$ \mathop{\lim}\nolimits_{t \to \infty}\tilde{\eta}_{i}(t)=0 $. 对于$ \dot{\bar{x}}_{i}(t)= (A_{i}-B_{i}K^{*}_{i})\bar{x}_{i}(t)+B_{i}L^{*}_{i}\tilde{\eta}(t) $, 由于$ A_{i}- B_{i}K^{*}_{i} $是赫尔维玆的, $ \mathop{\lim}\nolimits_{t \to \infty}\tilde{\eta}_{i}(t)=0 $, 根据文献[39]引理$ 9.1 $, 可知$ \mathop{\lim}\nolimits_{t \to \infty}\bar{x}_{i}(t) = 0 $. 根据式(4b)可知$ e_{i}(t)= C\bar{x}_{i}(t) $, 因此$ \mathop{\lim}\nolimits_{t \to \infty}e_{i}(t)=0 $, 实现了多智能体系统的量化通信下协同最优输出调节.

    在本节中, 我们将算法1应用于智能车联网的纵向协同自适应巡航控制[3, 40]. 协同自适应巡航控制是一种基于无线通信的智能自动驾驶策略, 车辆的通信拓扑如图 3所示, 外部系统仅可被车辆$ \#1 $直接访问.

    图 3  车辆通信拓扑图
    Fig. 3  Vehicular platoon communication topology

    利用以下模型对第$ i\;(i=1,\;2,\;3,\;4) $辆车进行建模:

    $$ \begin{split} x_{i}&=\upsilon_{i}\;\\ \dot{\upsilon}_{i}&=a_{i}\;\\ \dot{a}_{i}&=\sigma^{-1}_{i}a_{i}+\sigma^{-1}_{i}u_{i}+d_{i}\; \end{split} $$ (38)

    其中, $ x_{i} $, $ \upsilon_{i} $, $ a_{i} $, $ \sigma_{i} $分别为车辆$ \#i $发动机的位置、速度、加速度和时间常数. 常数$ d_{i} $是机械阻力与$ \sigma_{i} $和车辆$ \#i $质量的乘积之比. $ \sigma_{i} $与$ d_{i} $的值与文献[3]相同.

    车辆$ \#i $的参考轨迹$ x^{*}_{i} $和干扰信号$ d_{i} $均由以下外部系统产生

    $$ \begin{split}& \dot{\omega}_{1}=0.7\omega_{2}\;\\& \dot{\omega}_{2}=-0.7\omega_{1}\;\\& \dot{d_{i}}=d_{i}\omega_{2}\;\\& x^{*}_{i}=-5\omega_{1}-10(i+1)\omega_{2}\; \end{split} $$ (39)

    外部系统状态的初值为$ \omega(t)=[\omega_{1}(t)\; \; \; \omega_{2}(t)]= [0\; \; \; 1]^{{\mathrm{T}}} $.

    接下来, 对量化通信下的智能车联网系统进行仿真. 其中观测器参数$ \alpha=10 $, 调节函数$ s(k) $的初值为$ s(0)=0.05 $, 参数$ \mu=0.8 $. 外部系统状态估计误差$ \tilde{\eta}_{i}(t) $的收敛性如图 4所示.

    图 4  量化通信下外部系统状态估计误差$\tilde{\eta}_{i}(t)$的轨迹
    Fig. 4  The trajectory of the exosystem state estimation error $\tilde{\eta}_{i}(t)$ under quantized communication

    图 4可知, 选择的参数$ \alpha $能够保证$ \tilde{\eta}_{i}(t) $足够小, 当$ t>30 $s时, $ \tilde{\eta}_{i}(t)<10^{-6} $.

    当$ t<10 $s时, 我们应用初始控制策略$ u_{i0}= -K_{i0}x_{i}+\xi_{i} $, 其中探测噪声$ \xi_{i} $为不同频率的正弦信号的总和. 根据算法1迭代学习到控制增益$ K_{i,\;k} $和值$ P_{i,\;k} $, 其中每辆车的值$ P_{i,\;k} $与基于模型情况下得到的最优值$ P_{i}^{*} $的比较结果如图 5所示.

    图 5  每辆车$P_{i,\;k}$与最优解$P_{i}^{*}$的比较
    Fig. 5  Comparisons of $P_{i,\;k}$ and the optimal solution $ P_{i}^{*}$ of each vehicle

    图 5可知, 当迭代次数$ k=9 $时, $ P_{i,\;k} $能够收敛到最优解$ P_{i}^{*} $. 也就是说, 经过9次迭代之后, 所有车辆均能学习到最优控制值.

    当$ t=10 $s时, 通过学习到的最优控制增益$ (K_{i,\;k^*},\; P_{i,\;k^*}) $更新次优控制策略(3)并应用于智能车联网系统, 其实际轨迹$ x_{i} $与参考轨迹$ x^{*}_{i} $的跟踪情况如图 6所示. 仿真结果表明, 所有的车辆均能实现对参考轨迹的跟踪.

    图 6  智能互联自动驾驶车辆的实际轨迹$x_{i}$与参考轨迹$x^{*}_{i}$
    Fig. 6  Actual trajectories $x_{i}$ of connected and autonomous vehicles and their references $x^{*}_{i}$

    若当$ t=10 $s时, 不采用更新后的次优控制策略(3), 而是继续使用初始控制策略, 则初始控制策略控制下的智能车联网系统的实际轨迹$ x_{i} $与参考轨迹$ x^{*}_{i} $的跟踪情况如图 7所示. 从图 6图 7的对比可知, 通过算法1得到的次优控制策略能够实现车联网自动驾驶车辆在有干扰情况下对参考轨迹的跟踪.

    图 7  初始控制策略下智能互联自动驾驶车辆的实际轨迹$x_{i}$与参考轨迹$x^{*}_{i}$
    Fig. 7  Actual trajectories $x_{i}$ of intelligent connected and autonomous vehicles and their references $x^{*}_{i}$ under the initial control strategy

    接下来, 通过表 1比较量化通信对车辆间通信传输比特数的影响.

    表 1  达到$ ||P_{i,\;k}-P_{i}^{*}||<10^{-4} $有无量化通信传输的比特数
    Table 1  Transmitted bits with and without quantized communication to reach $ ||P_{i,\;k}-P_{i}^{*}||<10^{-4} $
    算法1下传输的比特数 无量化通信传输的比特数[3] 降低百分比
    80000 192000 58.33%
    下载: 导出CSV 
    | 显示表格

    表 1可知, 量化通信下只需要传输较少的比特数就能够达到特定的收敛误差, 量化通信下降低了$ 58.33\% $比特.

    本文研究量化通信下系统动态未知的连续时间多智能体系统的协同最优输出调节问题. 通过引入均匀量化器与编码−解码方案, 设计一种基于采样和量化数据的分布式协议, 用于观测外部系统状态, 在保证外部系统状态估计误差收敛的同时, 降低多智能体间的通信负担. 针对一类具有不确定系统动态的多智能体系统, 设计一种自适应动态规划方法, 用于多智能体系统的协同最优输出调节. 理论分析和在智能车联网自适应巡航控制系统上的仿真验证表明, 模型未知的多智能体系统能够在量化通信下实现渐近跟踪与干扰抑制. 我们未来的研究将考虑在有限带宽通信约束下, 针对外部系统状态与系统矩阵全部未知的非线性多智能体系统设计自适应最优控制策略.

  • 图  1  视频超分辨率重建数据集REDS (左)和Vimeo-90K (右)示例

    Fig.  1  Examples of video super-resolution datasets from REDS (left) and Vimeo-90K (right)

    图  2  部分VSR模型在REDS据集的可视化比较结果

    Fig.  2  Visual comparison results of VSR methods on REDS dataset

    图  3  部分VSR模型在Vid4数据集的可视化比较结果

    Fig.  3  Visual comparison results of VSR methods on Vid4 dataset

    图  4  本文的结构图

    Fig.  4  Architecture of the paper

    图  5  基于深度学习的视频超分辨率重建时间脉络图

    Fig.  5  Timeline of video super-resolution based on deep learning

    图  6  VSRNet结构图

    Fig.  6  Architecture of VSRNet

    图  7  VESPCN结构图

    Fig.  7  Architecture of VESPCN

    图  8  SOFVSR结构图

    Fig.  8  Architecture of SOFVSR

    图  9  TOFlow结构图

    Fig.  9  Architecture of TOFlow

    图  10  DUF结构

    Fig.  10  Architecture of DUF

    图  11  FSTRN结构图

    Fig.  11  Architecture of FSTRN

    图  12  TDAN结构图

    Fig.  12  Architecture of TDAN

    图  13  EDVR结构图

    Fig.  13  Architecture of EDVR

    图  14  TGA结构图

    Fig.  14  Architecture of TGA

    图  15  MuCAN结构图

    Fig.  15  Architecture of MuCAN

    图  16  MANA结构图

    Fig.  16  Architecture of MANA

    图  17  IAM结构图

    Fig.  17  Architecture of IAM

    图  18  VSR Transformer结构图

    Fig.  18  Architecture of VSR Transformer

    图  19  VRT结构图

    Fig.  19  Architecture of VRT

    图  20  DRVSR结构图

    Fig.  20  Architecture of DRVSR

    图  21  FRVSR结构图

    Fig.  21  Architecture of FRVSR

    图  22  RBPN结构图

    Fig.  22  Architecture of RBPN

    图  23  RLSP结构图

    Fig.  23  Architecture of RLSP

    图  24  RSDN结构图

    Fig.  24  Architecture of RSDN

    图  25  RRN结构图

    Fig.  25  Architecture of RRN

    图  26  DAP结构图

    Fig.  26  Architecture of DAP

    图  27  ETDM结构图

    Fig.  27  Architecture of ETDM

    图  28  TMP结构图

    Fig.  28  Architecture of TMP

    图  29  BRCN结构图

    Fig.  29  Architecture of BRCN

    图  30  RRCN结构图

    Fig.  30  Architecture of RRCN

    图  31  PFNL结构图和PFRB细节图

    Fig.  31  Architecture of PFNL and Detail of PFRB

    图  32  RISTN结构图

    Fig.  32  Architecture of RISTN

    图  33  LOVSR(左)和GOVSR(右)结构图

    Fig.  33  Architectures of LOVSR (left) and GOVSR (right)

    图  34  BasicVSR(左)和ICONVSR(右)结构图

    Fig.  34  Architectures of BasicVSR (left) and ICONVSR (right)

    图  35  TTVSR结构图

    Fig.  35  Architecture of TTVSR

    图  36  CTVSR结构图

    Fig.  36  Architecture of CTVSR

    图  37  RefVSR结构图

    Fig.  37  Architecture of RefVSR

    图  38  C2-Mathching结构图

    Fig.  38  Architecture of C2-Mathching

    图  39  RealBasicVSR结构图

    Fig.  39  Architecture of RealBasicVSR

    图  40  FTVSR结构图

    Fig.  40  Architecture of FTVSR

    图  41  BasicVSR++ 结构图

    Fig.  41  Architecture of BasicVSR++

    图  42  PSRT结构图

    Fig.  42  Architecture of PSRT

    图  43  IART结构图

    Fig.  43  Architecture of IART

    图  44  MFPI结构图

    Fig.  44  Architecture of MFPI

    图  45  DFVSR结构图

    Fig.  45  Architecture of DFVSR

    图  46  MIA-VSR结构图

    Fig.  46  Architecture of MIA-VSR

    图  47  RVRT结构图

    Fig.  47  Architecture of RVRT

    图  48  TeoGAN结构图

    Fig.  48  Architecture of TeoGAN

    图  49  StableVSR结构图

    Fig.  49  Architecture of StableVSR

    图  50  MGLD结构图

    Fig.  50  Architecture of MGLD

    图  51  Upscale-A-Video结构图

    Fig.  51  Architecture of Upscale-A-Video

    图  52  不同帧间对齐模式示意图

    Fig.  52  Illustration of different inter-frame alignment

    图  53  基于光流的显式运动对齐

    Fig.  53  Explicit alignment based on optical flow

    图  54  基于可变形卷积的对齐

    Fig.  54  Deformable convolution-based alignment

    图  55  光流引导的可变形对齐和光流引导的可变形注意力

    Fig.  55  Flow-guided deformable alignment and flow-guided deformable attention

    图  56  基于3D卷积的帧间对齐

    Fig.  56  Inter-frame alignment based on 3D convolution

    表  1  基于深度学习的视频超分辨率重建数据集

    Table  1  Datasets of video super-resolution based on deep learning

    数据集 类型 视频数量 帧数 分辨率 颜色空间
    合成数据集YUV25[15]训练集25386 $ \times $ 288YUV
    TDTFF[16]Turbine测试集5648 $ \times $ 528YUV
    Dancing950 $ \times $ 530
    Treadmill700$ \times $600
    Flag1000 $ \times $ 580
    Fan990 $ \times $ 740
    Vid4[13]Foliage测试集449720 $ \times $ 480RGB
    Walk47720 $ \times $ 480
    Calendar41720 $ \times $ 576
    City34704 $ \times $ 576
    YUV21[17]测试集21100352 $ \times $ 288YUV
    Venice[18]训练集11 0773 840$ \times $2 160RGB
    Myanmar[19]训练集15273 840$ \times $2 160RGB
    CDVL[20]训练集100301 920$ \times $1 080RGB
    UVGD[21]测试集163 840$ \times $2 160YUV
    LMT[22]训练集261 920$ \times $1 080YCbCr
    SPMCS[23]训练集和测试集97531960$ \times $540RGB
    MM542[24]训练集542321 280$ \times $720RGB
    UDM10[25]测试集10321 272$ \times $720RGB
    Vimeo-90K[12]训练集和测试集91 7017448$ \times $256RGB
    REDS[14]训练集和测试集2701001 280$ \times $720RGB
    Parkour[26]测试集14960$ \times $540RGB
    真实数据集RealVSR[27]训练集和测试集500501 024$ \times $512RGB/YCbCr
    VideoLQ[28]测试集501001 024$ \times $512RGB
    RealMCVSR[29]训练集和测试集1611 920$ \times $1 080RGB
    MVSR4$ \times $[30]训练集和测试集3001001 920$ \times $1 080RGB
    DTVIT[31]训练集和测试集1961001 920$ \times $1 080RGB
    YouHQ[32]训练集和测试集38 616321 920$ \times $1 080RGB
    下载: 导出CSV

    表  2  对双三次插值下采样后的视频进行VSR的性能对比结果

    Table  2  Performance comparison of video super-resolution algorithm with bicubic downsampling

    对比方法 训练帧数 参数量(M) 双三次插值下采样
    REDS (RGB通道) Vimeo-90K-T (Y通道) Vid4 (Y通道)
    Bicubic 26.14/0.7292 31.32/0.8684 23.78/0.6347
    VSRNet[40] 0.27 −/− −/− 22.81/0.6500
    VSRResFeatGAN[41] −/− −/− 24.50/0.7023
    VESPCN[42] −/− −/− 25.35/0.7577
    VSRResNet[41] −/− −/− 25.51/0.7530
    SPMC[23] 2.17 −/− −/− 25.52/0.7600
    3DSRNet[43] −/− −/− 25.71/0.7588
    RRCN[44] −/− −/− 25.86/0.7591
    TOFlow[12] 5/7 1.41 27.98/0.7990 33.08/0.9054 25.89/0.7651
    STARNet[45] 111.61 −/− 30.83/0.9290 −/−
    MEMC-Net[46] −/− 33.47/0.9470 24.37/0.8380
    STMN[47] −/− −/− 25.90/0.7878
    SOFVSR[48] 1.71 −/− −/− 26.01/0.7710
    RISTN[49] 3.67 −/− −/− 26.13/0.7920
    MMCNN[24] 10.58 −/− −/− 26.28/0.7844
    RTVSR[50] 15.00 −/− −/− 26.36/0.7900
    TDAN[51] 1.97 −/− −/− 26.42/0.7890
    D3DNet[52] −/7 2.58 −/− 35.65/0.9330 26.52/0.7990
    FFCVSR[53] −/− −/− 26.97/0.8300
    EVSRNet[54] 27.85/0.8000 −/− −/−
    StableVSR[55] 27.97/0.8000 −/− −/−
    DUF[56] 7/7 5.8 28.63/0.8251 −/− 27.33/0.8319
    PFNL[57] 7/7 3 29.63/0.8502 36.14/0.9363 26.73/0.8029
    DNSTNet[58] −/− 36.86/0.9387 27.21/0.8220
    RBPN[59] 7/7 12.2 30.09/0.8590 37.07/0.9435 27.12/0.8180
    DSMC[60] 11.58 30.29/0.8381 −/− 27.29/0.8403
    Boosted EDVR[31] 30.53/0.8699 −/− −/−
    TMP[61] 3.1 30.67/0.8710 −/− 27.10/0.8167
    MuCAN[62] 5/7 30.88/0.8750 37.32/0.9465 −/−
    MSFFN[63] −/− 37.33/0.9467 27.23/0.8218
    DAP[64] 15/5 30.59/0.8703 −/− −/−
    MultiBoot VSR[65] 60.86 31.00/0.8822 −/− −/−
    SSL-bi[66] 15/14 1.0 31.06/0.8933 36.82/0.9419 27.15/0.8208
    EDVR[67] 5/7 20.6 31.09/0.8800 37.61/0.9489 27.35/0.8264
    RLSP[68] 4.2 −/− 37.39/0.9470 27.15/0.8202
    TGA[69] 5.8 −/− 37.43/0.9480 27.19/0.8213
    KSNet-bi[70] 3.0 31.14/0.8862 37.54/0.9503 27.22/0.8245
    VSR-T[71] 5/7 32.6 31.19/0.8815 37.71/0.9494 27.36/0.8258
    PSRT-sliding[72] 5/− 14.8 31.32/0.8834 −/− −/−
    SeeClear[73] 5/5 229.23 31.32/0.8856 37.64/0.9503 27.80/0.8404
    DPR[74] 6.3 31.38/0.8907 37.11/0.9446 27.19/0.8243
    BasicVSR[75] 15/14 6.3 31.42/0.8909 37.18/0.9450 27.24/0.8251
    Boosted BasicVSR[31] 31.42/0.8917 −/− −/−
    SATeCo[76] 6/6 31.62/0.8932 −/− 27.44/0.8420
    IconVSR[75] 15/14 8.7 31.67/0.8948 37.47/0.9476 27.39/0.8279
    ICNet[77] 18.34 31.71/0.8963 37.72/0.9477 27.43/0.8287
    MSHPFNL[78] 7.77 −/− 36.75/0.9406 27.70/0.8472
    PA[79] 5/7 38.2 32.05/0.8941 −/− 28.02/0.8373
    FTVSR[80] 10.8 31.82/0.8960 −/− −/−
    $ C^2 $-Matching[81] 32.05/0.9010 −/− 28.87/0.8960
    ETDM[82] 8.4 32.15/0.9024 −/− −/−
    BasicVSR++[83] 30/14 7.3 32.39/0.9069 37.79/0.9500 27.79/0.8400
    RTA[84] 5/7 17 31.30/0.8850 37.84/0.9498 27.90/0.8380
    Semantic Lens[85] 5/− 31.42/0.8881 −/− −/−
    TCNet[86] 9.6 31.82/0.9002 37.94/0.9514 27.48/0.8380
    TTVSR[87] 50/− 6.8 32.12/0.9021 −/− −/−
    VRT[88] 16/7 35.6 32.19/0.9006 38.20/0.9530 27.93/0.8425
    CTVSR[89] 16/14 34.5 32.28/0.9047 −/− 28.03/0.8487
    FTVSR++[90] 10.8 32.42/0.9070 −/− −/−
    LGDFNet-BPP[91] 9.0 32.53/0.9007 −/− 27.99/0.8409
    PP-MSVSR-L[92] 7.4 32.53/0.9083 −/− −/−
    CFD-BasicVSR++[127] 30/7 7.5 32.51/0.9083 37.90/0.9504 27.84/0.8406
    RVRT[93] 30/14 10.8 32.75/0.9113 38.15/0.9527 27.99/0.8426
    DFVSR[94] 7.1 32.76/0.9081 38.25/0.9556 27.92/0.8427
    PSRT-recurrent[72] 16/14 13.4 32.72/0.9106 38.27/0.9536 28.07/0.8485
    MFPI[95] −/− 7.3 32.81/0.9106 38.28/0.9534 28.11/0.8481
    EvTexture[96] 15/− 8.9 32.79/0.9174 38.23/0.9544 29.51/0.8909
    MIA-VSR[97] 16/14 16.5 32.78/0.9220 38.22/0.9532 28.20/0.8507
    CFD-PSRT[127] 30/7 13.6 32.83/0.9140 38.33/0.9548 28.18/0.8503
    IART[98] 16/7 13.4 32.90/0.9138 38.14/0.9528 28.26/0.8517
    EvTexture+[96] 15/− 10.1 32.93/0.9195 38.32/0.9558 29.78/0.8983
    下载: 导出CSV

    表  3  对高斯模糊下采样后的视频进行VSR的性能对比结果

    Table  3  Performance comparison of video super-resolution algorithm with blur downsampling

    对比方法 训练帧数 参数量(M) 高斯模糊下采样
    UDM10 (Y通道) Vimeo-90K-T (Y通道) Vid4 (Y通道)
    Bicubic 28.47/0.8253 31.30/0.8687 21.80/0.5246
    BRCN[99] −/− −/− 24.43/0.6334
    ToFNet[12] 5/7 1.41 36.26/0.9438 34.62/0.9212 25.85/0.7659
    TecoGAN[100] 3.00 −/− −/− 25.89/−
    SOFVSR[48] 1.71 −/− −/− 26.19/0.7850
    RRN[101] 3.4 38.96/0.9644 −/− 27.69/0.8488
    TDAN[51] 1.97 −/− −/− 26.86/0.8140
    FRVSR[102] 5.1 −/− −/− 26.69/0.8220
    DUF[56] 7/7 5.8 38.48/0.9605 36.87/0.9447 27.38/0.8329
    RLSP[68] 4.2 38.48/0.9606 36.49/0.9403 27.48/0.8388
    PFNL[57] 7/7 3 38.74/0.9627 −/− 27.16/0.8355
    RBPN[59] 7/7 12.2 38.66/0.9596 37.20/0.9458 27.17/0.8205
    TMP[61] 3.1 −/− 37.33/0.9481 27.61/0.8428
    TGA[69] 5.8 38.74/0.9627 37.59/0.9516 27.63/0.8423
    SSL-bi[66] 15/14 1.0 39.35/0.9665 37.06/0.9458 27.56/0.8431
    RSDN[103] 6.19 −/− 37.23/0.9471 27.02/0.8505
    DAP[64] 15/5 39.50/0.9664 37.25/0.9472 −/−
    SeeClear[73] 5/5 229.23 39.72/0.9675 −/− −/−
    EDVR[67] 5/7 20.6 39.89/0.9686 37.81/0.9523 27.85/0.8503
    DPR[74] 6.3 39.72/0.9684 37.24/0.9461 27.89/0.8539
    BasicVSR[75] 15/14 6.3 39.96/0.9694 37.53/0.9498 27.96/0.8553
    IconVSR[75] 15/14 8.7 40.03/0.9694 37.84/0.9524 28.04/0.8570
    R2D2[104] 8.25 39.53/0.9670 −/− 28.13/0.9244
    FTVSR[80] 10.8 −/− −/− 28.31/0.8600
    FDAN[105] 39.91/0.9686 37.75/0.9522 27.88/0.8508
    PP-MSVSR[92] 1.45 40.06/0.9699 37.54/0.9499 28.13/0.8604
    GOVSR[106] 40.14/0.9713 37.63/0.9503 28.41/0.8724
    ETDM[82] 8.4 40.11/0.9707 −/− 28.81/0.8725
    TTVSR[87] 50/− 6.8 40.41/0.9712 37.92/0.9526 28.40/0.8643
    BasicVSR++[83] 30/14 7.3 40.72/0.9722 38.21/0.9550 29.04/0.8753
    CFD-BasicVSR++[127] 30/7 7.5 40.77/0.9726 38.36/0.9557 29.14/0.8760
    TCNet[86] 9.6 −/− −/− 28.44/0.8730
    VRT[88] 16/7 35.6 41.05/0.9737 38.72/0.9584 29.42/0.8795
    CTVSR[89] 16/14 34.5 41.20/0.9740 38.83/0.9580 29.28/0.8811
    FTVSR++[90] 10.8 −/− −/− 28.80/0.8680
    LGDFNet-BPP[91] 9.0 40.81/0.9756 −/− 29.39/0.8798
    RVRT[93] 30/14 10.8 40.90/0.9729 38.59/0.9576 29.54/0.8810
    DFVSR[94] 7.1 40.97/0.9733 38.51/0.9571 29.56/0.8983
    MFPI[95] −/− 7.3 41.08/0.9741 38.70/0.9579 29.34/0.8781
    下载: 导出CSV

    表  4  真实场景下的VSR性能对比结果

    Table  4  Performance comparison of real-world video super-resolution algorithm

    对比方法 推理帧数 RealVSR MVSR $ 4\times $
    PSNR/SSIM/LPIPS PSNR/SSIM/LPIPS
    RSDN[103] 之前帧 23.91/0.7743/0.224 23.15/0.7533/0.279
    FSTRN[107] 7 23.36/0.7683/0.240 22.66/0.7433/0.315
    TOF[12] 7 23.62/0.7739/0.220 22.80/0.7502/0.279
    TDAN[51] 7 23.71/0.7737/0.229 23.07/0.7492/0.282
    EDVR[67] 7 23.96/0.7781/0.216 23.51/0.7611/0.268
    BasicVSR[75] 所有帧 24.00/0.7801/0.209 23.38/0.7594/0.270
    MANA[108] 所有帧 23.89/0.7781/0.224 23.15/0.7513/0.285
    TTVSR[87] 所有帧 24.08/0.7837/0.213 23.60/0.7686/0.277
    ETDM[82] 所有帧 24.13/0.7896/0.206 23.61/0.7662/0.260
    BasicVSR++[83] 所有帧 24.24/0.7933/0.216 23.70/0.7713/0.263
    RealBasicVSR[28] 所有帧 23.74/0.7676/0.174 23.15/0.7603/0.202
    EAVSR[30] 所有帧 24.20/0.7862/0.208 23.61/0.7618/0.264
    EAVSR+[30] 所有帧 24.41/0.7953/0.212 23.94/0.7726/0.259
    EAVSRGAN+[30] 所有帧 23.99/0.7726/0.170 23.35/0.7611/0.199
    下载: 导出CSV

    表  5  不同帧间对齐方式的性能和参数比较

    Table  5  Performance and parameter comparisons of different inter-frame alignment

    对齐方式 参数量(M) 插值方法 光流
    GT SpyNet
    显式对齐(光流) 1.35 最近邻插值 31.84 31.78
    双线性插值 31.92 31.85
    双三次插值 31.93 31.89
    混合对齐(光流引导 1.60 双线性插值 32.08 31.98
    的可变形卷积)
    混合对齐(光流引导 1.56 双线性插值 32.03 31.94
    的可变形注意力)
    混合对齐(光流引导 1.35 最近邻插值 31.81 31.82'
    的图像块对齐)
    混合对齐(光流引导 1.36 基于注意力的 32.14 32.05
    的隐式对齐) 隐式插值
    下载: 导出CSV

    表  6  GeForce RTX 3090平台下VSR的性能和推理时间对比结果

    Table  6  Performance and inference time comparisons of VSR algorithm on GeForce RTX 3090 platform

    对比方法参数量(M)推理时间(ms)对齐方式双三次插值下采样高斯模糊下采样
    REDS (RGB通道)Vimeo-90K-T
    (Y通道)
    Vid4
    (Y通道)
    Vimeo-90K-T
    (Y通道)
    Vid4
    (Y通道)
    UDM10
    (Y通道)
    Bicubic<126.23/0.731931.32/0.868423.78/0.637431.30/0.868721.80/0.534628.47/0.8253
    TOFlow[12]1.41250显式27.96/0.798133.08/0.905425.89/0.765134.62/0.921225.85/0.765936.26/0.9438
    DUF[56]5.8737.5无需28.63/0.8251−/−27.33/0.831936.87/0.944727.38/0.832938.48/0.9605
    EDVR[67]20.6188.2隐式31.09/0.880037.61/0.948927.35/0.826437.81/0.952327.85/0.850339.89/0.9686
    TMP[61]3.131.5隐式30.67/0.8710−/−27.10/0.816737.33/0.948127.61/0.8428−/−
    BasicVSR[75]6.345.4显式31.42/0.890937.18/0.945027.24/0.825137.53/0.949827.96/0.855339.96/0.9694
    ICONVSR[75]8.758.4显式31.67/0.894837.47/0.947627.39/0.827937.84/0.952428.04/0.857040.03/0.9694
    TTVSR[87]6.8123.3混合32.12/0.9021−/−−/−37.92/0.952628.40/0.864340.41/0.9712
    VRT[88]35.61679混合32.17/0.900238.20/0.953027.93/0.842538.72/0.958429.37/0.879241.04/0.9737
    BasicVSR++[83]7.360.2混合32.39/0.906937.79/0.950027.79/0.840038.21/0.955029.04/0.875340.72/0.9722
    PSRT[72]13.41280.2混合32.72/0.910638.27/0.953628.07/0.8485−/−−/−−/−
    MIA-VSR[97]16.51194.6无需32.78 0.922038.22/0.953228.20/0.8507−/−−/−−/−
    下载: 导出CSV
  • [1] Wan Z, Zhang B, Chen D, et al. Bringing old films back to life. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 17694−17703
    [2] Li G, Ji J, Qin M, et al. Towards high-quality and efficient video super-resolution via spatial-temporal data overfitting. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver, Canada: IEEE, 2023. 10259−10269
    [3] Zhu H, Wei Y, Liang X, et al. CTP: Towards vision-language continual pretraining via compatible momentum contrast and topology preservation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Paris, France: IEEE, 2023. 22257−22267
    [4] Jiao S, Wei Y, Wang Y, et al. Learning mask-aware clip representations for zero-shot segmentation. In: Proceedings of the Annual Conference on Neural Information Processing Systems (NeurIPS). New Orleans, USA: 2023. 35631−35653.
    [5] Liu C, Sun D. On Bayesian adaptive video super resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(2): 346−360 doi: 10.1109/TPAMI.2013.127
    [6] Ma Z, Liao R, Tao X, et al. Handling motion blur in multi-frame super-resolution. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Boston, USA: IEEE, 2015. 5224−5232
    [7] Wu Y, Li F, Bai H, et al. Bridging component learning with degradation modelling for blind image super-resolution. IEEE Transactions on Multimedia, DOI: 10.1109/TMM.2022.3216115
    [8] 张帅勇, 刘美琴, 姚超, 林春雨, 赵耀. 分级特征反馈融合的深度图像超分辨率重建. 自动化学报, 2022, 48(4): 992−1003

    Zhang Shuai-Yong, Liu Mei-Qin, Yao Chao, Lin Chun-Yu, Zhao Yao. Hierarchical feature feedback network for depth super-resolution reconstruction. Acta Automatica Sinica, 2022, 48(4): 992−1003
    [9] Charbonnier P, Blanc-Feraud L, Aubert G, et al. Two deterministic half-quadratic regularization algorithms for computed imaging. In: Proceedings of 1st International Conference on Image Processing (ICIP). Austin, USA: IEEE, 1994. 168−172
    [10] Lai W S, Huang J B, Ahuja N, et al. Fast and accurate image super-resolution with deep laplacian pyramid networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 41(11): 2599−2613
    [11] Zha L, Yang Y, Lai Z, et al. A lightweight dense connected approach with attention on single image super-resolution. Electronics, 2021, 10(11): 1234 doi: 10.3390/electronics10111234
    [12] Xue T, Chen B, Wu J, et al. Video enhancement with task-oriented flow. International Journal of Computer Vision, 2019, 127(8): 1106−1125 doi: 10.1007/s11263-018-01144-2
    [13] Liu C, Sun D. A bayesian approach to adaptive video super resolution. In: Proceedings of IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Colorado Springs, USA: IEEE, 2011. 209−216
    [14] Nah S, Baik S, Hong S, et al. Ntire 2019 challenge on video deblurring and super-resolution: Dataset and study. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Long Beach, USA: IEEE, 2019. 1996−2005
    [15] Protter M, Elad M, Takeda H, et al. Generalizing the nonlocal-means to super-resolution reconstruction. IEEE Transactions on Image Processing, 2008, 18(1): 36−51
    [16] O. Shahar, A. Faktor, and M. Irani, Space-time super-resolution from a single video. In: Proceedings of the IEEE Conference Computer Vision and Pattern Recognition (CVPR). Colorado Springs, USA: IEEE, 2011. 3353−3360
    [17] Li D, Wang Z. Video superresolution via motion compensation and deep residual learning. IEEE Transactions on Computational Imaging, 2017, 3(4): 749−762 doi: 10.1109/TCI.2017.2671360
    [18] Venice[Online], available: https://www.harmonicinc.com/free-4k-demo-footage/, May 1, 2017
    [19] Myanmar 60p, Harmonic Inc. [Online], available: http://www.harmonicinc.com/resources/videos/4k-video-clip-center, May 1, 2017
    [20] ITS, "Consumer digital video library''[Online], available: https://www.cdvl.org, March 20, 2024
    [21] Mercat A, Viitanen M, Vanne J. UVG dataset: 50/120fps 4K sequences for video codec analysis and development. In: Proceedings of the ACM Multimedia Systems Conference. Istanbul, Turkey: ACM, 2020. 297−302
    [22] Liu D, Wang Z, Fan Y, et al. Robust video super-resolution with learned temporal dynamics. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV). Venice, Italy: IEEE, 2017. 2507−2515
    [23] Li D, Wang Z. Video superresolution via motion compensation and deep residual learning. IEEE Transactions on Computational Imaging, 2017, 3(4): 749−762 doi: 10.1109/TCI.2017.2671360
    [24] Wang Z, Yi P, Jiang K, et al. Multi-memory convolutional neural network for video super-resolution. IEEE Transactions on Image Processing, 2018, 28(5): 2530−2544
    [25] Yi P, Wang Z, Jiang K, et al. Progressive fusion video super-resolution network via exploiting non-local spatio-temporal correlations. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Seoul, Korea: IEEE, 2019. 3106−3115
    [26] Yu J, Liu J, Bo L, et al. Memory-augmented non-local attention for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 17834−17843
    [27] Yang X, Xiang W, Zeng H, et al. Real-world video super-resolution: A benchmark dataset and a decomposition based learning scheme. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Montreal, Canada: IEEE, 2021. 4781−4790
    [28] Chan K C K, Zhou S, Xu X, et al. Investigating tradeoffs in real-world video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 5962−5971
    [29] Lee J, Lee M, Cho S, et al. Reference-based video super-resolution using multi-camera video triplets. In: Proceedings of the IEEE/CVF conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 17824−17833
    [30] Wang R, Liu X, Zhang Z, et al. Benchmark dataset and effective inter-frame alignment for real-world video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver, Canada: IEEE, 2023. 1168−1177
    [31] Huang Y, Dong H, Pan J, et al. Boosting video super resolution with patch-based temporal redundancy optimization. In: Proceedings of International Conference on Artificial Neural Networks (ICANN). Heraklion, Greece: Springer, 2023. 362−375
    [32] Zhou S, Yang P, Wang J, et al. Upscale-A-Video: Temporal-consistent diffusion model for real-world video super-resolution. arXiv preprint arXiv: 2312.06640, 2023.
    [33] Wang X, Xie L, Dong C, et al. Real-esrgan: Training real-world blind super-resolution with pure synthetic data. In: Proceedings of the IEEE/CVF International Conference on Computer Vision Workshops (ICCVW). Montreal, Canada: IEEE, 2021. 1905−1914
    [34] Singh A, Singh J. Survey on single image based super-resolution—implementation challenges and solutions. Multimedia Tools and Applications, 2020, 79(3−5): 1641−1672
    [35] You Z, Li Z, Gu J, et al. Depicting beyond scores: Advancing image quality assessment through multi-modal language models. arXiv preprint arXiv: 2312.08962, 2023.
    [36] You Z, Gu J, Li Z, et al. Descriptive image quality assessment in the wild. arXiv preprint arXiv: 2405.18842, 2024.
    [37] Xie L, Wang X, Zhang H, et al. VFHQ: A high-quality dataset and benchmark for video face super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 657−666
    [38] Zhou F, Sheng W, Lu Z, et al. A database and model for the visual quality assessment of super-resolution videos. IEEE Transactions on Broadcasting, 2024, 70(2): 516−532 doi: 10.1109/TBC.2024.3382949
    [39] Jin J, Zhang X, Fu X, et al. Just noticeable difference for deep machine vision. IEEE Transactions on Circuits and Systems for Video Technology, 2021, 32(6): 3452−3461
    [40] Kappeler A, Yoo S, Dai Q, et al. Video super-resolution with convolutional neural networks. IEEE Transactions on Computational Imaging, 2016, 2(2): 109−122 doi: 10.1109/TCI.2016.2532323
    [41] Lucas A, Lopez-Tapia S, Molina R, et al. Generative adversarial networks and perceptual losses for video super-resolution. IEEE Transactions on Image Processing, 2019, 28(7): 3312−3327 doi: 10.1109/TIP.2019.2895768
    [42] Caballero J, Ledig C, Aitken A, et al. Real-time video super-resolution with spatio-temporal networks and motion compensation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, USA: IEEE, 2017. 4778−4787
    [43] Kim S Y, Lim J, Na T, et al. 3DSRNet: video super-resolution using 3D convolutional neural networks. arXiv preprint arXiv: 1812.09079, 2018.
    [44] Li D, Liu Y, Wang Z. Video super-resolution using non-simultaneous fully recurrent convolutional network. IEEE Transactions on Image Processing, 2018, 28(3): 1342−1355
    [45] Haris M, Shakhnarovich G, Ukita N. Space-time-aware multi-resolution video enhancement. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020. 2859−2868
    [46] Bao W, Lai W S, Zhang X, et al. MEMC-Net: Motion estimation and motion compensation driven neural network for video interpolation and enhancement. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019, 43(3): 933−948
    [47] Zhu X, Li Z, Lou J, et al. Video super-resolution based on a spatio-temporal matching network. Pattern Recognition, 2021, 110: 107619 doi: 10.1016/j.patcog.2020.107619
    [48] Wang L, Guo Y, Liu L, et al. Deep video super-resolution using HR optical flow estimation. IEEE Transactions on Image Processing, 2020, 29: 4323−4336 doi: 10.1109/TIP.2020.2967596
    [49] Zhu X, Li Z, Zhang X Y, et al. Residual invertible spatio-temporal network for video super-resolution. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Honolulu, USA: AAAI, 2019. 5981−5988
    [50] Bare B, Yan B, Ma C, et al. Real-time video super-resolution via motion convolution kernel estimation. Neurocomputing, 2019, 367: 236−245 doi: 10.1016/j.neucom.2019.07.089
    [51] Tian Y, Zhang Y, Fu Y, et al. TDAN: Temporally-deformable alignment network for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020. 3360−3369
    [52] Ying X, Wang L, Wang Y, et al. Deformable 3D convolution for video super-resolution. IEEE Signal Processing Letters, 2020, 27: 1500−1504 doi: 10.1109/LSP.2020.3013518
    [53] Yan B, Lin C, Tan W. Frame and feature-context video super-resolution. In: Proceedings of the 33th AAAI Conference on Artificial Intelligence (AAAI). Honolulu, USA: AAAI, 2019: 5597−5604
    [54] Liu S, Zheng C, Lu K, et al. Evsrnet: Efficient video super-resolution with neural architecture search. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Nashville, USA: IEEE, 2021. 2480−2485
    [55] Rota C, Buzzelli M, van de Weijer J. Enhancing perceptual quality in video super-resolution through temporally-consistent detail synthesis using diffusion models. arXiv preprint arXiv: 2311.15908, 2023.
    [56] Jo Y, Oh S W, Kang J, et al. Deep video super-resolution network using dynamic upsampling filters without explicit motion compensation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Salt Lake City, USA: IEEE, 2018. 3224−3232
    [57] Yi P, Wang Z, Jiang K, et al. Progressive fusion video super-resolution network via exploiting non-local spatio-temporal correlations. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Seoul, Korea: IEEE, 2019. 3106−3115
    [58] Sun W, Sun J, Zhu Y, et al. Video super-resolution via dense non-local spatial-temporal convolutional network. Neurocomputing, 2020, 403: 1−12 doi: 10.1016/j.neucom.2020.04.039
    [59] Haris M, Shakhnarovich G, Ukita N. Recurrent back-projection network for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 3897−3906
    [60] Liu H, Zhao P, Ruan Z, et al. Large motion video super-resolution with dual subnet and multi-stage communicated upsampling. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Virtual Event: AAAI, 2021. 2127−2135
    [61] Zhang Z, Li R, Guo S, et al. TMP: Temporal motion propagation for online Video super-sesolution. arXiv preprint arXiv: 2312.09909, 2023.
    [62] Li W, Tao X, Guo T, et al. Mucan: Multi-correspondence aggregation network for video super-resolution. In: Proceedings of the European Conference on Computer Vision (ECCV). Glasgow, UK: Springer, 2020. 335−351
    [63] Song H, Xu W, Liu D, et al. Multi-stage feature fusion network for video super-resolution. IEEE Transactions on Image Processing, 2021, 30: 2923−2934 doi: 10.1109/TIP.2021.3056868
    [64] Fuoli D, Danelljan M, Timofte R, et al. Fast online video super-resolution with deformable attention pyramid. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). Waikoloa, USA: IEEE, 2023. 1735−1744
    [65] Kalarot R, Porikli F. Multiboot VSR: Multi-stage multi-reference bootstrapping for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Long Beach, USA: IEEE, 2019. 2060−2069
    [66] Xia B, He J, Zhang Y, et al. Structured sparsity learning for efficient video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2023: 22638−22647
    [67] Wang X, Chan K C K, Yu K, et al. EDVR: Video restoration with enhanced deformable convolutional networks. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Long Beach, USA: IEEE, 2019. 1954−1963
    [68] Fuoli D, Gu S, Timofte R. Efficient video super-resolution through recurrent latent space propagation. In: Proceedings of the IEEE/CVF International Conference on Computer Vision Workshop (ICCVW). Seoul, Korea: IEEE, 2019. 3476−3485
    [69] Isobe T, Li S, Jia X, et al. Video super-resolution with temporal group attention. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020. 8008−8017
    [70] Jin S, Liu M, Yao C, et al. Kernel Dimension Matters: To activate available kernels for real-time video super-resolution. In: Proceedings of the ACM International Conference on Multimedia (ACM MM). Ottawa, Canada: ACM, 2023. 8617−8625
    [71] Cao J, Li Y, Zhang K, et al. Video super-resolution transformer. arXiv preprint arXiv: 2106.06847, 2021.
    [72] Shi S, Gu J, Xie L, et al. Rethinking alignment in video super-resolution transformers. In: Proceedings of the Annual Conference on Neural Information Processing Systems (NeurIPS). New Orleans, USA: 2022. 36081−36093
    [73] Tang Q, Zhao Y, Liu M, et al. SeeClear: Semantic distillation enhances pixel condensation for video super-resolution. arXiv preprint arXiv: 2410.05799, 2024.
    [74] Huang C, Li J, Chu L, et al. Disentangle propagation and restoration for efficient video recovery. In: Proceedings of the ACM International Conference on Multimedia (ACM MM). Ottawa, Canada: ACM, 2023. 8336−8345
    [75] Chan K C K, Wang X, Yu K, et al. Basicvsr: The search for essential components in video super-resolution and beyond. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Nashville, USA: IEEE, 2021. 4947−4956
    [76] Chen Z, Long F, Qiu Z, et al. Learning spatial adaptation and temporal coherence in diffusion models for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2024. 9232−9241
    [77] Leng J, Wang J, Gao X, et al. Icnet: Joint alignment and reconstruction via iterative collaboration for video super-resolution. In: Proceedings of the ACM International Conference on Multimedia (ACM MM). Lisboa, Portugal: ACM, 2022. 6675−6684
    [78] Yi P, Wang Z, Jiang K, et al. A progressive fusion generative adversarial network for realistic and consistent video super-resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 44(5): 2264−2280
    [79] Zhang F, Chen G, Wang H, et al. Multi-scale video super-resolution transformer with polynomial approximation. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33(9): 4496−4506 doi: 10.1109/TCSVT.2023.3278131
    [80] Qiu Z, Yang H, Fu J, et al. Learning spatiotemporal frequency-transformer for compressed video super-resolution. In: Proceedings of the European Conference on Computer Vision (ECCV). Tel Aviv, Israel: Springer, 2022. 257−273
    [81] Jiang Y, Chan K C K, Wang X, et al. Reference-based image and video super-resolution via C2-matching. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2022, 45(7): 8874−8887
    [82] Isobe T, Jia X, Tao X, et al. Look back and forth: Video super-resolution with explicit temporal difference modeling. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 17411−17420
    [83] Chan K C K, Wang X, Yu K, et al. Basicvsr: The search for essential components in video super-resolution and beyond. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Nashville, USA: IEEE, 2021. 4947−4956
    [84] Zhou K, Li W, Lu L, et al. Revisiting temporal alignment for video restoration. In: Proceedings/CVF of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 6053−6062
    [85] Tang Q, Zhao Y, Liu M, et al. Semantic lens: Instance-centric semantic alignment for video super-resolution. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Vancouver, Canada: AAAI, 2024. 5154−5161
    [86] Liu M, Jin S, Yao C, et al. Temporal consistency learning of inter-frames for video super-resolution. IEEE Transactions on Circuits and Systems for Video Technology, 2022, 33(4): 1507−1520
    [87] Liu C, Yang H, Fu J, et al. Learning trajectory-aware transformer for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 5687−5696
    [88] Liang J, Cao J, Fan Y, et al. VRT: A video restoration transformer. IEEE Transactions on Image Processing, 2024, 33: 2171−2182 doi: 10.1109/TIP.2024.3372454
    [89] Tang J, Lu C, Liu Z, et al. CTVSR: Collaborative spatial-temporal transformer for video super-resolution. IEEE Transactions on Circuits and Systems for Video Technology, DOI: 10.1109/TCSVT.2023.3340439
    [90] Qiu Z, Yang H, Fu J, et al. Learning degradation-robust spatiotemporal frequency-transformer for video super-resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(12): 14888−14904 doi: 10.1109/TPAMI.2023.3312166
    [91] Zhang C, Wang X, Xiong R, et al. Local-global dynamic filtering network for video super-resolution. IEEE Transactions on Computational Imaging, 2023, 9: 963−976 doi: 10.1109/TCI.2023.3321980
    [92] Jiang L, Wang N, Dang Q, et al. PP-MSVSR: multi-stage video super-resolution. arXiv preprint arXiv: 2112.02828, 2021.
    [93] Liang J, Fan Y, Xiang X, et al. Recurrent video restoration transformer with guided deformable attention. Proceedings of the Annual Conference on Neural Information Processing Systems (NeurIPS). New Orleans, USA: 2022. 378−393
    [94] Dong S, Lu F, Wu Z, et al. DFVSR: directional frequency video super-resolution via asymmetric and enhancement alignment network. In: Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI). Macao, China: IJCAI, 2023. 681−689
    [95] Li F, Zhang L, Liu Z, et al. Multi-frequency representation enhancement with privilege information for video super-resolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Paris, France: IEEE, 2023. 12814−12825
    [96] Kai D, Lu J, Zhang Y, et al. EvTexture: Event-driven texture enhancement for video super-resolution. arXiv preprint arXiv: 2406.13457, 2024.
    [97] Zhou X, Zhang L, Zhao X, et al. Video Super-Resolution Transformer with Masked Inter&Intra-Frame Attention. arXiv preprint arXiv: 2401.06312, 2024.
    [98] Xu K, Yu Z, Wang X, et al. An implicit alignment for video super-resolution. arXiv preprint arXiv: 2305.00163, 2023.
    [99] Huang Y, Wang W, Wang L. Video super-resolution via bidirectional recurrent convolutional networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 40(4): 1015−1028
    [100] Chu M, Xie Y, Mayer J, et al. Learning temporal coherence via self-supervision for GAN-based video generation. ACM Transactions on Graphics, 2020, 39(4): 75
    [101] Isobe T, Zhu F, Jia X, et al. Revisiting temporal modeling for video super-resolution. arXiv preprint arXiv: 2008.05765, 2020.
    [102] Sajjadi M S M, Vemulapalli R, Brown M. Frame-recurrent video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Salt Lake City, USA: IEEE, 2018. 6626−6634
    [103] Isobe T, Jia X, Gu S, et al. Video super-resolution with recurrent structure-detail network. In: Proceedings of the European Conference on Computer Vision (ECCV). Glasgow, UK: Springer, 2020. 645−660
    [104] Baniya A A, Lee T K, Eklund P W, et al. Online video super-resolution using information replenishing unidirectional recurrent model. Neurocomputing, 2023, 546: 126355 doi: 10.1016/j.neucom.2023.126355
    [105] Lin J, Huang Y, Wang L. FDAN: Flow-guided deformable alignment network for video super-resolution. arXiv preprint arXiv: 2105.05640, 2021.
    [106] Yi P, Wang Z, Jiang K, et al. Omniscient video super-resolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Montreal, Canada: IEEE, 2021. 4429−4438
    [107] Li S, He F, Du B, et al. Fast spatio-temporal residual network for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 10522−10531
    [108] Yu J, Liu J, Bo L, et al. Memory-augmented non-local attention for video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 17834−17843
    [109] Tao X, Gao H, Liao R, et al. Detail-revealing deep video super-resolution. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV). Venice, Italy: IEEE, 2017. 4472−4480
    [110] Yang X, He C, Ma J, et al. Motion-guided latent diffusion for temporally consistent real-world video super-resolution. arXiv preprint arXiv: 2312.00853, 2023.
    [111] Liu H, Ruan Z, Zhao P, et al. Video super-resolution based on deep learning: a comprehensive survey. Artificial Intelligence Review, 2022, 55(8): 5981−6035 doi: 10.1007/s10462-022-10147-y
    [112] Tu Z, Li H, Xie W, et al. Optical flow for video super-resolution: A survey. Artificial Intelligence Review, 2022, 55(8): 6505−6546 doi: 10.1007/s10462-022-10159-8
    [113] Baniya A A, Lee G, Eklund P, et al. A methodical study of deep learning based video super-resolution. Authorea Preprints, DOI: 10.36227/techrxiv.23896986.v1
    [114] 江俊君, 程豪, 李震宇, 刘贤明, 王中元. 深度学习视频超分辨率技术概述. 中国图象图形学报, 2023, 28(7): 1927−1964 doi: 10.11834/jig.220130

    Jiang Jun-Jun, Cheng Hao, Li Zhen-Yu, Liu Xian-Ming, Wang Zhong-Yuan. Deep learning based video-related super-resolution technique: A survey. Journal of Image and Graphics, 2023, 28(7): 1927−1964 doi: 10.11834/jig.220130
    [115] Dong C, Loy C C, He K, et al. Image super-resolution using deep convolutional networks. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2015, 38(2): 295−307
    [116] Drulea M, Nedevschi S. Total variation regularization of local-global optical flow. In: Proceedings of the International IEEE Conference on Intelligent Transportation Systems (ITSC). Washington, USA: IEEE, 2011. 318−323
    [117] Haris M, Shakhnarovich G, Ukita N. Deep back-projection networks for super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Salt Lake City, USA: IEEE, 2018. 1664−1673
    [118] Dai J, Qi H, Xiong Y, et al. Deformable convolutional networks. In: Proceedings of the IEEE International Conference on Computer Vision (ICCV). Venice, Italy: IEEE, 2017. 764−773
    [119] Zhu X, Hu H, Lin S, et al. Deformable convnets v2: More deformable, better results. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Long Beach, USA: IEEE, 2019. 9308−9316
    [120] Chan K C K, Wang X, Yu K, et al. Understanding deformable alignment in video super-resolution. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Virtual Event: AAAI, 2021: 973−981
    [121] Butler D J, Wulff J, Stanley G B, et al. A naturalistic open source movie for optical flow evaluation. In: Proceedings of European Conference on Computer Vision (ECCV). Florence, Italy: Springer, 2012. 611−625
    [122] Lian W, Lian W. Sliding window recurrent network for efficient video super-resolution. In: Proceedings of the European Conference on Computer Vision Workshops (ECCVW). Tel Aviv, Israel: Springer Nature Switzerland, 2022. 591−601
    [123] Xiao J, Jiang X, Zheng N, et al. Online video super-resolution with convolutional kernel bypass grafts. IEEE Transactions on Multimedia, 2023, 25: 8972−8987 doi: 10.1109/TMM.2023.3243615
    [124] Li D, Shi X, Zhang Y, et al. A simple baseline for video restoration with grouped spatial-temporal shift. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver, Canada: IEEE, 2023. 9822−9832
    [125] Geng Z, Liang L, Ding T, et al. Rstt: Real-time spatial temporal transformer for space-time video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, USA: IEEE, 2022. 17441−17451
    [126] Lin L, Wang X, Qi Z, et al. Accelerating the training of video super-resolution models. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Washington, USA: AAAI, 2023. 1595−1603
    [127] Li H, Chen X, Dong J, et al. Collaborative feedback discriminative propagation for video super-resolution. arXiv preprint arXiv: 2404.04745, 2024.
    [128] Hu M, Jiang K, Wang Z, et al. Cycmunet+: Cycle-projected mutual learning for spatial-temporal video super-resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023, 45(11): 13376−13392
    [129] Xiao Y, Yuan Q, Jiang K, et al. Local-global temporal difference learning for satellite video super-resolution. IEEE Transactions on Circuits and Systems for Video Technology, 2024, 34(4): 2789−2802 doi: 10.1109/TCSVT.2023.3312321
    [130] Hui Y, Liu Y, Liu Y, et al. VJT: A video transformer on joint tasks of deblurring, low-light enhancement and denoising. arXiv preprint arXiv: 2401.14754, 2024.
    [131] Song Y, Wang M, Yang Z, et al. NegVSR: Augmenting negatives for generalized noise modeling in real-world video super-resolution. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Vancouver, Canada: AAAI, 2024. 10705−10713
    [132] Wang Y, Isobe T, Jia X, et al. Compression-aware video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Vancouver, Canada: IEEE, 2023. 2012−2021
    [133] Youk G, Oh J, Kim M. FMA-Net: Flow-guided dynamic filtering and iterative feature refinement with multi-attention for joint video super-resolution and deblurring. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2024. 44−55
    [134] Zhang Y, Yao A. RealViformer: Investigating attention for real-world video super-resolution. arXiv preprint arXiv: 2407.13987, 2024.
    [135] Xiang X, Tian Y, Zhang Y, et al. Zooming slow-mo: Fast and accurate one-stage space-time video super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2020: 3370−3379
    [136] Jeelani M, Cheema N, Illgner-Fehns K, et al. Expanding synthetic real-world degradations for blind video super resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW). Vancouver, Canada: IEEE, 2023. 1199−1208
    [137] Bai H, Pan J. Self-supervised deep blind video super-resolution. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024, 46(7): 4641−4653 doi: 10.1109/TPAMI.2024.3361168
    [138] Pan J, Bai H, Dong J, et al. Deep blind video super-resolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (CVPR). Seattle, USA: IEEE, 2024. 4811−4820
    [139] Chen H, Li W, Gu J, et al. Low-res leads the way: Improving generalization for super-resolution by self-supervised learning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2024. 25857−25867
    [140] Yuan J, Ma J, Wang B, et al. Content-decoupled contrastive learning-based implicit degradation modeling for blind image super-resolution. arXiv preprint arXiv: 2408.05440, 2024.
    [141] Chen Y H, Chen S C, Lin Y Y, et al. MoTIF: Learning motion trajectories with local implicit neural functions for continuous space-time video super-resolution. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV). Paris, France: IEEE, 2023. 23131−23141
    [142] Huang C, Li J, Chu L, et al. Arbitrary-scale video super-resolution guided by dynamic context. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Vancouver, Canada: AAAI, 2024. 2294−2302
    [143] Li Z, Liu H, Shang F, et al. SAVSR: Arbitrary-scale video super-resolution via a learned scale-adaptive network. In: Proceedings of the AAAI Conference on Artificial Intelligence (AAAI). Vancouver, Canada: AAAI, 2024. 3288−3296
    [144] Huang Z, Huang A, Hu X, et al. Scale-adaptive feature aggregation for efficient space-time video super-resolution. In: Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). Waikoloa, USA: IEEE, 2024. 4228−4239
    [145] Xu Y, Park T, Zhang R, et al. VideoGigaGAN: Towards detail-rich video super-resolution. arXiv preprint arXiv: 2404.12388, 2024.
    [146] He Q, Wang S, Liu T, et al. Enhancing measurement precision for rotor vibration displacement via a progressive video super resolution network. IEEE Transactions on Instrumentation and Measurement, 2024, 73: 1−13
    [147] Chang J, Zhao Z, Jia C, et al. Conceptual compression via deep structure and texture synthesis. IEEE Transactions on Image Processing, 2022, 31: 2809−2823 doi: 10.1109/TIP.2022.3159477
    [148] Chang J, Zhang J, Li J, et al. Semantic-aware visual decomposition for image coding. International Journal of Computer Vision, 2023, 131(9): 2333−2355 doi: 10.1007/s11263-023-01809-7
    [149] Ren B, Li Y, Liang J, et al. Sharing key semantics in transformer makes efficient image restoration. arXiv preprint arXiv: 2405.20008, 2024.
    [150] Wu R, Sun L, Ma Z, et al. One-step effective diffusion network for real-world image super-resolution. arXiv preprint arXiv: 2406.08177, 2024.
    [151] Sun H, Li W, Liu J, et al. Coser: Bridging image and language for cognitive super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2024. 25868−25878
    [152] Wu R, Yang T, Sun L, et al. Seesr: Towards semantics-aware real-world image super-resolution. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, USA: IEEE, 2024. 25456−25467
    [153] Zhang Y, Zhang H, Chai X, et al. MRIR: Integrating multimodal insights for diffusion-based realistic image restoration. arXiv preprint arXiv: 2407.03635, 2024.
    [154] Zhang Y, Zhang H, Chai X, et al. Diff-restorer: Unleashing visual prompts for diffusion-based universal image restoration. arXiv preprint arXiv: 2407.03636, 2024.
    [155] Ouyang H, Wang Q, Xiao Y, et al. Codef: Content deformation fields for temporally consistent video processing. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). 2024: 8089−8099
    [156] Hu J, Gu J, Yu S, et al. Interpreting low-level vision models with causal effect maps. arXiv preprint arXiv: 2407.19789, 2024.
    [157] Gu J, Dong C. Interpreting super-resolution networks with local attribution maps. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Virtual: IEEE, 2021. 9199−9208
    [158] Cao J, Liang J, Zhang K, et al. Towards interpretable video super-resolution via alternating optimization. In: Proceedings of the European Conference on Computer Vision (ECCV). Tel Aviv, Israel: Springer, 2022. 393−411
  • 加载中
计量
  • 文章访问数:  147
  • HTML全文浏览量:  234
  • 被引次数: 0
出版历程
  • 收稿日期:  2024-04-29
  • 录用日期:  2024-10-16
  • 网络出版日期:  2025-03-06

目录

/

返回文章
返回