之前很早就想试着做一下试着把顶会的论文浏览一遍看一下自己感兴趣的，顺便统计一下国内高校或者研究机构的研究方向，下面是作为一个图像处理初学者在浏览完论文后的 觉得有趣的文章：
ICCV2017 论文浏览记录
1.google deepmind :Look, Listen and Learn 多信息融合感觉很厉害
2.The Weizmann Institute of Science：Non-Uniform Blind Deblurring by Reblurring 非均匀盲模糊
3.中科大(微软亚洲研究院) :CVAE-GAN: Fine-Grained Image Generation through Asymmetric Training 对抗网络对物体分类
4.University of California：Exploiting Spatial Structure for Localizing Manipulated Image Regions CNN+LSTM 图像处理
5.CVLab, EPFL, Lausanne, Switzerland：Deep Occlusion Reasoning for Multi-Camera Multi-Target Detection 对比RCNN有优势
6.Computer Vision Lab, TU Dresden ：Bounding Boxes, Segmentations and Object Coordinates:How Important is Recognition for 3D Scene Flow Estimation in Autonomous Driving Scenarios 无人驾驶，车的检测
7.University of Pennsylvania：Am I a Baller? Basketball Performance Assessment
from First-Person Videos 第一视角来评估篮球运动员的表现
8.Institute of Geodesy and Photogrammetry, ETH Zurich（苏黎世理工）：Semantically Informed Multiview Surface Refinement 图像语义分割
9.University of Maryland：Soft-NMS – Improving Object Detection With One Line of Code（听名字好像很牛逼）
10.Michigan State University：Illuminating Pedestrians via Simultaneous Detection & Segmentation 行人检测分割
11.The University of Nottingham:Binarized Convolutional Landmark Localizers for Human Pose Estimation and Face Alignment with Limited Resources 人体检测
12.University of Surrey Guildford:SubUNets: End-to-end Hand Shape and Continuous Sign Language Recognition CNN+LSTM端到端的手语识别
13.CMU：One Network to Solve Them All — Solving Linear Inverse Problems using Deep Projection Models 图像修复
14.南理工（南大）：Adversarial PoseNet: A Structure-aware Convolutional Network for Human Pose Estimation 人姿态估计
15.Intel Labs：Fast Image Processing with Fully-Convolutional Networks，图片的后期处理，爱摄像的可以看
16.北航：Look, Perceive and Segment: Finding the Salient Objects in Images
via Two-stream Fixation-Semantic CNNs SOD目标检测
17.港中文（腾训优图）：Makeup-Go: Blind Reversion of Portrait Edit 照片去美化，人像还原
18.University of Southern California：Query-guided Regression Network with Context Policy for Phrase Grounding 图像描述
19.国立清华(微软亚洲研究院):Show, Adapt and Tell: Adversarial Training of Cross-domain Image Captioner 图像语义理解
20.清华：Surface Normals in the Wild 表面方向单目图像的深度估计
21.清华：SegFlow: Joint Learning for Video Object Segmentation and Optical Flow 视频分割 光流算法
22. 中科大（港中文）：Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal Attention Mechanism 目标跟踪
23.Institute of Mathematics of the Romanian Academy：Unsupervised learning from video to detect foreground objects in single images目标检测
24.Google Brain：Pixel Recursive Super Resolution 超分辨率
25.Microsoft Research Asia：Deformable Convolutional Networks
26.港中文：Towards Diverse and Natural Image Descriptions via a Conditional GAN 对抗网络对图像的理解
27Georgia Institute of Technology：Learning Cooperative Visual Dialog Agents with Deep Reinforcement Learning 协作对话
28.Imperial College London：Semantic Image Synthesis via Adversarial Learning 用自然语言描述图像合成
29.Inria：BlitzNet: A Real-Time Deep Network for Scene Understanding场景分割 贝叶斯网络
30.CMU：Cut, Paste and Learn: Surprisingly Easy Synthesis for Instance Detection 小数据集对物体进行识别
31.美图：Parallel Tracking and Verifying: A Framework for Real-Time and High Accuracy Visual Tracking 跟踪 蛮好的
32.Graz University of Technology: Detect to Track and Track to Detect 目标检测与跟踪（给了github 进去却是空的，有点可惜）
33.MPI for Intelligent Systems：Semantic Video CNNs through Representation Warping视频语义分析
34.CMU：Learning Background-Aware Correlation Filters for Visual Tracking 跟踪
35.CMU：Need for Speed: A Benchmark for Higher Frame Rate Object Tracking 高速跟踪的基准
36.中科大：A Multimodal Deep Regression Bayesian Network for Affective Video Content Analyses 深度回归贝叶斯的视频情感分析
37.清华：VQS: Linking Segmentations to Questions and Answers for Supervised Attention in VQA and Question-Focused Semantic Segmentation 图像分割+语义理解
38.University of Texas at Austin：On-Demand Learning for Deep Image Restoration 图像复原
39.University of Southern California：TALL: Temporal Activity Localization via Language Query 通过语言查询进行时态活动定位
40.CS Department Stanford University：Fine-grained Recognition in the Wild: A Multi-Task Domain Adaptation Approach 野外物体识别 李飞飞的
41.Nanyang Technological University:An Empirical Study of Language CNN for Image Captioning 图像语义
42.天大：Learning Dynamic Siamese Network for Visual Object Tracking 追踪
43.University of Illinois, Urbana-Champaign：Aligned Image-Word Representations Improve Inductive Transfer Across
Vision-Language Tasks 视觉问答
44.Stanford University:Characterizing and Improving Stability in Neural Style Transfer 画风转换 李飞飞
45.University Politehnica of Bucharest：Unsupervised object segmentation in video by selection of highly probable positive features 图像分割
46.香港大学：SCNet: Learning Semantic Correspondence语义对应
47.西交大：Channel Pruning for Accelerating Very Deep Neural Networks 修剪深网络 孙剑
48.Facebook AI Research：Mask R-CNN 好像性能可以和faster RCNN媲美 VIP
49.UC Berkeley：Localizing Moments in Video with Natural Language 视频语义
50.The University of Tokyo:Joint Detection and Recounting of Abnormal Events by Learning Deep Generic Knowledge 图像语义
51.University of California：Learning to Reason: End-to-End Module Networks for Visual Question Answering 视觉问答
52.北大：Centered Weight Normalization in Accelerating Training of Deep Neural Networks 对网络权重进行重新规划
53.CMU：Learning Policies for Adaptive Tracking with Deep Feature Cascades 跟踪
54.国科大：Wavelet-SRNet: A Wavelet-based CNN for Multi-scale Face Super Resolution 人脸超分辨率
55.University of California：Scene Parsing with Global Context Embedding 图像分割
56.ETH Zurich：DSLR-Quality Photos on Mobile Devices with Deep Convolutional Networks 学习提高照片的质量
57NUS Graduate School for Integrative Science and Engineering：Video Scene Parsing with Predictive Feature Learning
视频情景解析
58.Stanford University：Inferring and Executing Programs for Visual Reasoning 情景推理 李飞飞
59.Rochester Institute of Technology：An Analysis of Visual Question Answering Algorithms 视觉问答
60.Joint learning of object and action detectors 物体跟踪与动作识别
61.Tel Aviv University：Temporal Tessellation: A Unified Approach for Video Analysis 视频语义分析
62.University of Mannheim：Higher-Order Minimum Cost Lifted Multicuts for Motion Segmentation物体分割
63.Korea University：CDTS: Collaborative Detection, Tracking, and Segmentation for Online Multiple Object Segmentation in Videos 视频检测分割
64.Twitter：Fast Face-swap Using Convolutional Neural Networks 快速换脸，好像蛮有意思的
65.Stanford University：Dense-Captioning Events in Videos 视频语义 能上下文 李飞飞
66.National University of Singapore：Dual-Glance Model for Deciphering Social Relationships 图片中的社会关系 在图像语义理解时可能有用
67.港中文：Identity-Aware Textual-Visual Matching with Latent Co-attention 图片与句子描述之间的相识度
68.大连理工：Is Second-order Information Helpful for Large-scale Visual Recognition? 关于卷积神经网深层信息
69.Snap Inc（阅后即焚应用），雅虎，谷歌：Learning from Noisy Labels with Distillation 从杂乱标签中学习，对标注不精确
70.港中文：Learning to Disambiguate by Asking Discriminative Questions 消除回答奇异的方法，关于VQA
71.成电：Leveraging Weak Semantic Relevance for Complex Video Event Classification 视频事件分析
72.港中文：Scene Graph Generation from Objects, Phrases and Region Captions 图像语义
73.港中文：Situation Recognition with Graph Neural Networks 识别情景
74.清华：Recurrent Topic-Transition GAN for Visual Paragraph Generation 对抗网络来进行语义生成
75.深圳大学（港中文）：Cascaded Feature Network for Semantic Segmentation of RGB-D Images 图像语义分割
76.港中文（商汤）：HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis 人姿态分析
77.商汤（港中文）：Recurrent Scale Approximation for Object Detection in CNN 目标的多尺度检测 汤晓鸥
78.中科院：Referring Expression Generation and Comprehension via Attributes 通过索引产生语义
79.Facebook AI Research：Predicting Deeper into the Future of Semantic Segmentation 预测语义
80.中山大学：Deep Dual Learning for Semantic Image Segmentation 语义理解减少对标签数据的依赖
81.University of Illinois at Urbana-Champaign:Recurrent Models for Situation Recognition 情景识别
82.University of Zurich：Rotation equivariant vector field networks cnn+旋转变化
83.IIT Hyderabad： Attentive Semantic Video Generation using Captions 视频字幕 语义
84.Bosch：Universal Adversarial Perturbations Against Semantic Image Segmentation 加噪声对图像分割
85.POSTECH：MarioQA: Answering Questions by Watching Gameplay Videos 通过游戏视频进行QA
86.IIIT-Delhi：Face Sketch Matching via Coupled Deep Transform Learning 人脸与素描的匹配
87.University of Maryland：SSH: Single Stage Headless Face Detector 数脸
88.Yonsei University：Modelling the Scene Dependent Imaging in Cameras with a Deep Neural Network 用深度学习改进照片的质量
89.Mapillary Research：The Mapillary Vistas Dataset for Semantic Understanding of Street Scenes街头情景语义理解
90.阿里（西电，西交）：Hierarchical Multimodal LSTM for Dense Visual-Semantic Embedding 语义理解
91.MIT CSAIL：Personalized Cinemagraphs using Semantic Understanding and Collaborative Learning 语义理解和协作学习
92.University of Southern California：Realistic Dynamic Facial Textures from a Single Image using GANs用对抗网络进行面部纹理处理
93.Max Planck Institute for Informatics Saarland Informatics Campus Saabrucken：Towards a Visual Privacy Advisor:
Understanding and Predicting Privacy Risks in Images图像关于隐私
94.Aristotle University of Thessaloniki Thessalonik：Learning Bag-of-Features Pooling for Deep Convolutional Neural Networks 提出一种新的卷积BoF
95.Weakly-supervised learning of visual relations 视觉关联弱监督学习，和语义有部分相关
96.University of Illinois at Urbana-Champaign：Phrase Localization and Visual Relationship Detection with Comprehensive Image-Language Cues 图像语义
97.IMEC, Belgium：Encoder Based Lifelong Learning 终身学习 好像蛮有意思的
98.Technische Universitat Munich ：Learning in an Uncertain World: Representing Ambiguity Through Multiple Hypotheses 预测不确定 表达
99.Max Planck Institute for Intelligent Systems: EnhanceNet: Single Image Super-Resolution Through Automated Texture Synthesis 超分辨率 纹理
100.University of Maryland：Guided Perturbations: Self-corrective Behavior in Convolutional Neural Networks 主动向卷积网络中加入扰动来提高网络稳定性
101.Georgia Institute of Technology：Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localization网络可视化
国内：
102.Saarland Informatics Campus:Speaking the Same Language: Matching Machine to Human Captions by Adversarial Training 用对抗网络将视频与字幕对应起来
103.CMU：What Actions are Needed for Understanding Human Actions in Videos? 视频人行为分析
104.University of Jena：Generalized orderless pooling performs implicit salient matching CNN中对pooling层进行优化
105.Rutgers University, Department of Electrical and Computer Engineering：Generating High-Quality Crowd Density Maps using Contextual Pyramid CNNs 数人头
106.University of Massachusetts：Reasoning about Fine-grained Attribute Phrases using Reference Games 提出框架描述图片的细微区别
107.Google Research:Revisiting Unreasonable Effectiveness of Data in Deep Learning Era 蛮有意思的
108.港中文（Uber，旷视，腾讯优图）：Detail-revealing Deep Video Super-resolution 视频超分辨率 贾佳亚
109.上交：SORT: Second-Order Response Transform for Visual Recognition 将二阶操作引入神经网络
110.CMU：Transitive Invariance for Self-supervised Visual Representation Learning 何凯明
111.Uppsala University：Neural Ctrl-F: Segmentation-free Query-by-String Word Spotting in Handwritten Manuscript Collections 单词位置识别
112.Facebook AI Research：Unsupervised Creation of Parameterized Avatars 根据人脸画卡通图像
113.The Johns Hopkins University：Adversarial Examples for Semantic Segmentation and Object Detection语义分割与目标检测
114.The Johns Hopkins University：Genetic CNN 遗传CNN
115. 港中文：Supplementary Meta-Learning: Towards a Dynamic Model For Deep Neural Networks 补充单元学习
116. 港中文（深圳大学）：Online Robust Image Alignment via Subspace Learning from Gradient Orientations
117.Berkeley AI Research (BAIR) laboratory：Unpaired Image-to-Image Translation
using Cycle-Consistent Adversarial Networks 喔
118.The University of Adelaide：Towards Context-aware Interaction Recognition for Visual Relationship Detection语义
119.Athens Technological & Educational Inst:Offline Handwritten Signature Modeling and Verification Based on Archetypal Analysis 签名识别
国内：
1.华科（微软亚洲研究院）：Ensemble Diffusion for Retrieval
华科：Transformed Low -rank Model for Line Pattern Noise Removal 去噪
华科：AOD-Net: All-in-One Dehazing Network
华科（西北工业）：When Unsupervised Domain Adaptation Meets Tensor Representations
华科（西北大学）：Jointly Attentive Spatial-Temporal Pooling Networks for Video-based Person Re-Identification 人的检测
华科：Multi-modal Factorized Bilinear Pooling with Co-Attention Learning for Visual Question Answering 视觉问答
2.中科大(微软亚洲研究院) :CVAE-GAN: Fine-Grained Image Generation through
Asymmetric Training
中科大（微软亚洲研究院）：Coherent Online Video Style Transfer 关于视频画风转换
中科大：A Multimodal Deep Regression Bayesian Network for Affective Video Content Analyses 深度回归贝叶斯的视频情感分析
中科大（港中文）：Online Multi-Object Tracking Using CNN-based Single Object Tracker with Spatial-Temporal
中科大：DualNet: Learn Complementary Features for Image Recognition 图像识别
中科大：VegFru: A Domain-Specific Dataset for Fine-grained Visual Categorization VegFru分类可视化
中科大：Deep Facial Action Unit Recognition from Partially Labeled Data
中科大（中山大学）：Boosting Image Captioning with Attributes 图像的字幕
中科大：Learning Multi-Attention Convolutional Neural Network for Fine-Grained Image Recognition 图像识别
中科大（微软）：Flow-Guided Feature Aggregation for Video Object Detection 目标检测
3.华南理工（兰州大学）：A Joint Intrinsic-Extrinsic Prior Model for Retinex
华南理工（香港城市大学，中科院）Delving into Salient Object Subitizing and Detection 图像分割
4.香港理工（哈工大）：Higher-order Integration of Hierarchical Convolutional Activations for Fine-grained Visual Categorization
香港理工（西交大，哈工大）：Joint Convolutional Analysis and Synthesis Sparse Representation for Single Image Layer Separation
香港理工（深圳大学）：3D Surface Detail Enhancement from A Single Normal Map 3D重建
香港理工：Towards More Accurate Iris Recognition Using Deeply Learned Spatially Corresponding Features 虹膜
香港理工（西电）：Multi-channel Weighted Nuclear Norm Minimization for Real Color Image Denoising 去噪
5.中科院（国科大，南审）：Egocentric Gesture Recognition Using Recurrent 3D Convolutional Neural Networks with Spatiotemporal Transformer Modules 手势识别；
中科院（国科大）：Deep Adaptive Image Clustering深自适应聚类
中科院：Attention Mechanism 目标跟踪
中科院：Deep Direct Regression for Multi-Oriented Scene Text Detection 文本识别
中科院：Beyond Face Rotation: Global and Local Perception GAN for Photorealistic and Identity Preserving Frontal View Synthesis关于人脸
中科院（上海科技大学，华为）：Learning Discriminative Latent Attributes for Zero-Shot Classification（迁移学习没看过）
中科院（西北工业）：Image2song: Song Retrieval via Bridging Image Content and Lyric Words 根据图片内容选歌词
中科院：Referring Expression Generation and Comprehension via Attributes 通过索引产生语义
中科院（西交大）：Tensor RPCA by Bayesian CP Factorization with Complex Noise 视频分割，不同场景场景进行处理
中科院（腾讯AILab ，南理工，国科大）：Video Deblurring via Semantic Segmentation and Pixel-Wise Non-Linear Kernel
中科院（国科大）：Depth and Image Restoration from Light Field in a Scattering Medium 图像恢复
中科院：Recursive Spatial Transformer (ReST) for Alignment-Free Face Recognition 人脸识别
中科院（天大，南方科技大学，腾讯AILab，港中文）：Range Loss for Deep Face Recognition with Long-Tailed Training Data 人脸识别相关
中科院：S3FD: Single Shot Scale-invariant Face Detector 人脸检测器
中科院（360，国科大）：Scale-adaptive Convolutions for Scene Parsing 情景解析
中科院（南审）：CoupleNet: Coupling Global Structure with Local Parts for Object Detection 目标检测
6.清华：HashNet: Deep Learning to Hash by Continuation
清华：Surface Normals in the Wild 单目图像的深度估计
清华 ：SegFlow: Joint Learning for Video Object Segmentation and Optical Flow视频分割 光流算法
清华：VQS: Linking Segmentations to Questions and Answers for Supervised Attention in VQA and Question-Focused Semantic Segmentation 图像分割+语义理解
清华（腾讯AILab，360AI institute）FoveaNet: Perspective-aware Urban Scene Parsing 城市场景识别 自动驾驶
清华：Recurrent Topic-Transition GAN for Visual Paragraph Generation 对抗网络来进行语义生成
清华：Mutual Enhancement for Detection of Multiple Logos in Sports Videos 商标检测
清华：Cross-Modal Deep Variational Hashing 哈希多媒体检索
清华：Learning Efficient Convolutional Networks through Network Slimming 加速网络
清华：Decoder Network over Lightweight Reconstructed Feature for Fast Semantic Style Transfer 风格转化
清华：Revisiting Cross-channel Information Transfer for Chromatic Aberration Correction 通过多通道关联来矫正色差
清华（南理工）：Learning to Super-Resolve Blurry Face and Text Images 超分辨率
清华：Single Image Action Recognition using Semantic Body Part Actions 动作识别
7.深圳大学（西北工业，华南理工）：A Self-Balanced Min-Cut Algorithm for Image Clustering 聚类
深圳大学（港中文）：Cascaded Feature Network for Semantic Segmentation of RGB-D Images 图像语义分割
8.南理工（南大）：Adversarial PoseNet: A Structure-aware Convolutional Network for Human Pose Estimation 人姿态估计（图像理解）
南理工（大连理工，腾讯优图，UC）：Learning Discriminative Data Fitting Functions for Blind Image Deblurring 图像去模糊
9.北航：Look, Perceive and Segment: Finding the Salient Objects in Images via Two-stream Fixation-Semantic CNNs SOD目标检测
北航：Primary Video Object Segmentation via Complementary CNNs and Neighborhood Reversible Flow 视频图像分割
北航：Embedding 3D Geometric Features for Rigid Object Part Segmentation 三维分割
北航（清华）：BodyFusion: Real-time Capture of Human Motion and Surface Geometry Using a Single Depth Camera 运动捕捉 表情捕捉
10.北京交通大学：Low-Rank Tensor Completion: A Pseudo-Bayesian Learning Approach
北京交通大学（中科院）：Robust Object Tracking based on Temporal and Spatial Deep Networks 基于时间和空间上的目标跟踪
11.港中文（腾讯优图）：Makeup-Go: Blind Reversion of Portrait Edit 照片去美化，人像还原
港中文：Towards Diverse and Natural Image Descriptions via a Conditional GAN 对抗网络对图像的理解
港中文：Identity-Aware Textual-Visual Matching with Latent Co-attention 图片与句子描述之间的相识度
港中文：Learning to Disambiguate by Asking Discriminative Questions 消除回答奇异的方法，关于VQA
港中文：Scene Graph Generation from Objects, Phrases and Region Captions 图像语义
港中文：Situation Recognition with Graph Neural Networks 识别情景
港中文：SGN: Sequential Grouping Networks for Instance Segmentation SGN图像分割
港中文：Chained Cascade Network for Object Detection 连级网络对对象的检测
港中文：Video Frame Synthesis using Deep Voxel Flow 视频帧合成 汤晓鸥
港中文（商汤）：Learning Deep Neural Networks for Vehicle Re-ID with Visual-spatio-temporal Path Proposals车辆再识别
港中文（Uber，旷视，腾讯优图）：Detail-revealing Deep Video Super-resolution 视频超分辨率 贾佳亚
港中文（腾讯优图，旷视）：Zero-order Reverse Filtering反向滤波 贾佳亚
港中文：Learning Feature Pyramids for Human Pose Estimation 人姿态
港中文：StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks
港中文：Supplementary Meta-Learning: Towards a Dynamic Model For Deep Neural Networks 补充单元学习
港中文：Temporal Action Detection with Structured Segment Networks 视频分享 汤晓鸥
港中文（深圳大学）：Online Robust Image Alignment via Subspace Learning from Gradient Orientations
港中文：Unsupervised Learning of Stereo Matching 无监督学习立体匹配
港中文：Be Your Own Prada: Fashion Synthesis with Structural Coherence 关于时尚
港中文（商汤）：HydraPlus-Net: Attentive Deep Features for Pedestrian Analysis 人姿态分析
12.国立清华:No More Discrimination: Cross City Adaptation of Road Scene Segmenters 图像分割（无人驾驶）
国立清华(微软亚洲研究院):Show, Adapt and Tell: Adversarial Training of Cross-domain Image Captioner 图像语义理解
国立清华：Robust Pseudo Random Fields for Light-Field Stereo Matching经验的贝叶斯定理框架-鲁棒伪随机场
国立清华：Anticipating Daily Intention using On-Wrist Motion Triggered Sensing
13.海康（复旦，上交）：Focusing Attention: Towards Accurate Text Recognition in Natural Images OCR
14.腾讯优图：Weakly- and Self-Supervised Learning for Content-Aware Deep Image Retargeting 选择合适的放大方式对图像进行缩放
腾讯优图（港中文）：High-Quality Correspondence and Segmentation Estimation for Dual-Lens Smart-Phone Portraits 手机双摄 图像处理
腾讯优图：Weakly Supervised Object Localization Using Things and Stuff Transfer
15.复旦：Temporal Context Network for Activity Localization in Videos
复旦（腾讯AILab）：Multi-scale Deep Learning Architectures for Person Re-identification人的
复旦（清华）：DSOD: Learning Deeply Supervised Object Detectors from Scratch 目标检测器
复旦：Towards 3D Human Pose Estimation in the Wild: a Weakly-supervised Approach人姿态检测
16.西电：Deep Clustering via Joint Convolutional Autoencoder Embedding and Relative Entropy Minimization
17.大连理工（南理工，桂林电子）：Blind Image Deblurring with Outlier Handling 模糊图像清晰
大连理工，哈工大：Is Second-order Information Helpful for Large-scale Visual Recognition? 关于卷积神经网深层信息
大连理工：Stepwise Metric Promotion for Unsupervised Video Person Re-identification 无监督学习人识别
大连理工：Unsupervised Domain Adaptation for Face Recognition in Unlabeled Videos 无监督学习人脸识别
大连理工：A Stagewise Refinement Model for Detecting Salient Objects in Images 分割
大连理工：Amulet: Aggregating Multi-level Convolutional Features for Salient Object Detection 突出对象检测
大连理工：Learning Uncertain Convolutional Features for Accurate Saliency Detection特征点
大连理工（首都师范大学）：Surface Registration via Foliation
18.国科大（中科院，港中文）：RPAN: An End-to-End Recurrent Pose-Attention Network for Action
Recognition in Videos
国科大：Wavelet-SRNet: A Wavelet-based CNN for Multi-scale Face Super Resolution 人脸超分辨率
国科大（中科院）：Multimodal Gaussian Process Latent Variable Models with Harmonization 高斯过程隐变量模型
国科大（微软亚洲研究院，中科院）：Human Pose Estimation using Global and Local Normalization 用局部和全局来判断人的姿态
国科大：Soft Proposal Networks for Weakly Supervised Object Localization 定位
19.山东大学(微软研究院):A Generic Deep Architecture for Single Image Reflection Removal and Image Smoothing 去镜子反光
20.西交大：Complex Event Detection by Identifying Reliable Shots from Untrimmed Videos
西交大：Channel Pruning for Accelerating Very Deep Neural Networks 修剪深网络 孙剑
西交大：Monocular 3D Human Pose Estimation by Predicting Depth on Joints 单目人姿态估计并预测深度
西交大：Predicting Human Activities Using Stochastic Grammar 人行动的预测
西交大：Should We Encode Rain Streaks in Video as Deterministic or Stochastic?
西交大：View Adaptive Recurrent Neural Networks for High Performance Human Action Recognition from Skeleton Data 人骨骼
21.美图：Parallel Tracking and Verifying: A Framework for Real-Time and High Accuracy Visual Tracking 跟踪 蛮好的
22.南开：Structure-measure: A New Way to Evaluate Foreground Maps 目标分割 程明明实验室的的
23.上交（腾讯优图）：RMPE: Regional Multi-Person Pose Estimation多人姿态估计
上交（北大）：Performance Guaranteed Network Acceleration via High-Order Residual Quantization网络加速
上交（上海科学技术大学，上海大学）：Semi-Global Weighted Least Squares in Image Filtering 滤波
上交：SORT: Second-Order Response Transform for Visual Recognition 将二阶操作引入神经网络
20.西北工业（华南理工）：Self-paced Kernel Estimation for Robust Blind Im age Deblurring 图像复原
西北工业：Monocular Dense 3D Reconstruction of a Complex Dynamic Scene from Two Perspective Frames
西北工业：Towards End-to-end Text Spotting with Convolutional Recurrent Neural Networks文字识别
西北工业：Efficient Global 2D-3D Matching for Camera Localization in a Large-Scale 3D Map 地图定位
西北工业：Supervision by Fusion: Towards Unsupervised Learning of Deep Salient Object Detector 图像分割
21.天大：Learning Dynamic Siamese Network for Visual Object Tracking 追踪
22.香港大学：High-Resolution Shape Completion Using Deep Neural Networks for Global Structure and Local Geometry Inference 模型复原
香港大学：SCNet: Learning Semantic Correspondence语义对应
23.台湾大学：Drone-based Object Counting by Spatially Regularized Regional Proposal Network
24.北大：Centered Weight Normalization in Accelerating Training of Deep Neural Networks 对网络权重进行重新规划
北大（图森）：Factorized Bilinear Models for Image Recognition 双线性分类
北大：Attention-aware Deep Reinforcement Learning for Video Face Recognition 视频人脸识别
北大（北理工）：Learning long-term dependencies for action recognition with a biologically-inspired deep network 提出生物启发网络
北大（北理工）：Exploiting Multi-Grain Ranking Constraints for Precisely Searching Visually-similar Vehicles 检测测量相识度
北大（上海交通大学）：Hard-Aware Deeply Cascaded Embedding
25.百度：WordSup: Exploiting Word Annotations for Character based Text Detection 文字识别
26.香港科技大学（清华）：SurfaceNet: An End-to-end 3D Neural Network for Multiview Stereopsis
香港科技大学（上海交通大学）：Online Video Object Detection using Association LSTM 用LSTM进行视频人检测
香港科技大学（华南理工）Lattice Long Short-Term Memory for Human Action Recognition 用CNN和LSTM来记录人的动作
香港科技大学：Temporal Dynamic Graph LSTM for Action-driven Video Object Detection LSTM 事件的检测
27.腾讯AILab （360AI Institute）：Video Scene Parsing with Predictive Feature Learning 视频情景解析
腾讯AILab（商汤）：Detecting Faces Using Inside Cascaded Contextual CNN 联级CNN 检测
28.成电：Leveraging Weak Semantic Relevance for Complex Video Event Classification 视频事件分析
29.国立台湾：Unrolled Memory Inner-Products: An Abstract GPU Operator for Efficient Vision-Related Computations 关于GPU
30.北师大：3DCNN-DQN-RNN: A Deep Reinforcement Learning Framework for Semantic Parsing of Large-scale 3D Point Clouds 大规模深度学习点云
31.浙大：Infant Footprint Recognition 婴儿足迹识别
浙大（微软研究院）：In this paper, we address the problem of person reidentification, which refers to associating the persons captured from different cameras 人识别
32.合肥工业（西南交通，腾讯AILab，360AI Institute）：Neural Person Search Machines 找人
33.商汤（港中文）：Recurrent Scale Approximation for Object Detection in CNN 目标的多尺度检测 汤晓鸥
商汤（ 清华，港中文）：Orientation Invariant Feature Embedding and Spatial Temporal Regularization for Vehicle Re-identification 车辆识别
34.上海科技大学：A Revisit of Sparse Coding Based Anomaly Detection in Stacked RNN Framework 稀疏编码
上海科技大学：Catadioptric HyperSpectral Light Field Imaging 成像技术
上海科技大学：Deep Free-Form Deformation Network for Object-Mask Registration 目标检测
上海科技大学：Generative Modeling of Audible Shapes for Object Perception 图像与发声物体识别
上海科技大学：Ray Space Features for Plenoptic Structure-from-Motion
上海科技大学：Structured Attentions for Visual Question Answering
35.中山大学：Deep Dual Learning for Semantic Image Segmentation 语义理解减少对标签数据的依赖
中山大学：Deep Growing Learning 半监督学习相关的
中山大学（商汤）：Multi-label Image Recognition by Recurrently Discovering Attentional Regions 多标签识别
中山大学：RGB-Infrared Cross-Modality Person Re-Identification RGB+红外 人的识别
中山大学：Cross-view Asymmetric Metric Learning for Unsupervised Person Re-identification 人检测
36.南京大学（上交）：ThiNet: A Filter Level Pruning Method for Deep Neural Network Compression 关于神经网络的修剪
南京大学：Adaptive Feeding: Achieving Fast and Accurate Detections by Adaptively Combining Object Detectors 目标检测
南京大学（清华）：Monocular Free-head 3D Gaze Tracking with Deep Learning and Geometry Constraints 目光检测
37.北邮：Object-level Proposals 图像边缘相关的
38.香港城市大学：Least Squares Generative Adversarial Networks
香港城市大学（深网视界）：CREST: Convolutional Residual Learning for Visual Tracking 视觉跟踪
39.阿里（西电，西交）：Hierarchical Multimodal LSTM for Dense Visual-Semantic Embedding 语义理解
40.上海大学：Multi-stage Multi-recursive-input Fully Convolutional Networks for Neuronal Boundary Detection 卷积网络进行边缘检测
41.同济：Compositional Human Pose Regression 用回归来判断人的姿态
同济：Scale Recovery for Monocular Visual Odometry Using Depth Estimated with
Deep Convolutional Neural Fields 单目深度估计
42.哈工大：Non-Rigid Object Tracking via Deformable Patches using Shape-Preserved KC and Level Sets 提出了一种滤波器来对对象进行跟踪
43.帝视技术：Image Super-Resolution Using Dense Skip Connections 超分辨率
44.北理工：Deep Cropping via Attention Box Prediction and Aesthetics Assessment
北理工：Super-Trajectory for Video Segmentation 视频分割
北理工：Transferring Objects: Joint Inference of Container and Human Pose 人姿势识别
45.百度：Deep Metric Learning with Angular Loss 特征检测方法
46.360：Recurrent 3D-2D Dual Learning for Large-pose Facial Landmark Detection
47.台湾中央研究院（国立台湾大学）：DeepCD: Learning Deep Complementary Descriptors for Patch Representation 共同学习 图像补丁
48.厦门大学：We propose a deep network architecture for the pansharpening problem called PanNet 关于PanNet
49.香港浸会大学：Dynamic Label Graph Matching for Unsupervised Video Re-Identification 标签估计
50.青岛大学（大连理工）：Intrinsic 3D Dynamic Surface Tracking based on Dynamic Ricci Flow and Teichmuller Map 3D
51.云南师范大学：Point Set Registration with Global-local Correspondence and Transformation Estimation
52.武汉科技大学（美图）：Saliency Pattern Detection by Ranking Structured Trees
统计了后明白了：
你爸爸永远是你爸爸 ，港中文的论文数给跪了
