ECCV 2020 论文汇总（注意力模型、事件相机、知识蒸馏、图像去雾去雨去噪、图像超分辨率、语义分割，等方向）

ECCV 2020 论文汇总

（注意力模型、事件相机、知识蒸馏、图像去雾去雨去噪、图像超分辨率、语义分割，等方向）

         [oral, spotlight]

[Post]

[Dehazing]

[Enhancement]

[Dinoising]

[Deraining]

[Event]

[Attention]

[Super-Resolution]

[Wavelet]

[Bayesian]

[Semantic Segmentation]

[Distillation]

---

---

[oral, spotlight]

• AiR: Attention with Reasoning Capability
• Invertible Image Rescaling
• End-to-End Object Detection with Transformers
• Rewriting a Deep Generative Model
• Conditional Convolutions for Instance Segmentation
• Content-Aware Unsupervised Deep Homography Estimation
• Gradient Centralization: A New Optimization Technique for Deep Neural Networks
• Learning Stereo from Single Images
• Diffraction Line Imaging
• Semantic Flow for Fast and Accurate Scene Parsing
• Self-Challenging Improves Cross-Domain Generalization
• Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulation
• Orientation-aware Vehicle Re-identification with Semantics-guided Part Attention Network
• Mining Cross-Image Semantics for Weakly Supervised Semantic Segmentation
• Towards Streaming Image Understanding
• ForkGAN: Seeing into the Rainy Night
• TopoGAN: A Topology-Aware Generative Adversarial Network
• ExchNet: A Unified Hashing Network for Large-Scale Fine-Grained Image Retrieval
•  

[Post]

• Convolutional Occupancy Networks
• Circumventing Outliers of AutoAugment with Knowledge Distillation
• TuiGAN: Learning Versatile Image-to-Image Translation with Two Unpaired Images
• GAN Slimming: All-in-One Unified GAN Compression
• Binarized Neural Network for Single Image Super Resolution
• Single-Image Depth Prediction Makes Feature Matching Easier
•  

[Dehazing]

• Nighttime Defogging Using High-Low Frequency Decomposition and Grayscale-Color Networks
• HardGAN: A Haze-Aware Representation Distillation GAN for Single Image Dehazing
• Physics-based Feature Dehazing Networks
•  

[Enhancement]

• PieNet: Personalized Image Enhancement Network
• Multi-level Wavelet-based Generative Adversarial Network for Perceptual Quality Enhancement of Compressed Video
• Global and Local Enhancement Networks For Paired and Unpaired Image Enhancement
• Learning Enriched Features for Real Image Restoration and Enhancement
• Early Exit Or Not: Resource-Efficient Blind Quality Enhancement for Compressed Images
• Low light video Enhancement using Synthetic Data Produced with an Intermediate Domain Mapping
• URIE: Universal Image Enhancement for Visual Recognition in the Wild
• Dynamic Low-light Imaging with Quanta Image Sensors
•  

[Dinoising]

• Reconstructing the Noise Manifold for Image Denoising
• Burst Denoising via Temporally Shifted Wavelet Transforms
• Robust and On-the-fly Dataset Denoising for Image Classification
• Learning Graph-Convolutional Representations for Point Cloud Denoising
• Stochastic Frequency Masking to Improve Super-Resolution and Denoising Networks
• Spatial-Adaptive Network for Single Image Denoising
• Unpaired Learning of Deep Blind Image Denoising
• Spatial Hierarchy Aware Residual Pyramid Network for Time-of-Flight Depth Denoising
• A Decoupled Learning Scheme for Real-world Burst Denoising from Raw Images
• Practical Deep Raw Image Denoising on Mobile Devices
•  

[Deraining]

• Beyond Monocular Deraining: Paired Rain Removal Networks via Unpaired Semantic Understanding
• Rethinking Image Deraining via Rain Streaks and Vapors
• ForkGAN: Seeing into the Rainy Night
•  

[Event]

• A Differentiable Recurrent Surface for Asynchronous Event-Based Data
• Learning Modality Interaction for Temporal Sentence Localization and Event Captioning in Videos
• Traffic Accident Analysis by Cause and Effect Events Localization
• How to Train Your Event Camera Neural Network [paper]
• Entropy Minimisation Framework for Event-based Vision Model Estimation
• Event-based Asynchronous Sparse Convolutional Networks
• REMIND Your Neural Network to Prevent Catastrophic Forgetting
• Learning Event-Driven Video Deblurring and Interpolation
• Learning to See in the Dark with Events
• CLAWS: Clustering Assisted Weakly Supervised Learning with Normalcy Suppression for Anomalous Event Detection
• Stereo Event-based Particle Tracking Velocimetry for 3D Fluid Flow Reconstruction
• Globally-Optimal Event Camera Motion Estimation
• Jointly learning visual motion and confidence from local patches in event cameras
• Spike-FlowNet: Event-based Optical Flow Estimation with Energy-Efficient Hybrid Neural Networks
• Event Enhanced High-Quality Image Recovery
• RhyRNN: Rhythmic RNN for Recognizing Events in Long and Complex Videos
•  

[Attention]

• Example-Guided Image Synthesis across Arbitrary Scenes using Masked Spatial-Channel Attention and Self-Supervision
• Unsupervised Domain Attention Adaptation Network for Caricature Attribute Recognition
• History Repeats Itself: Human Motion Prediction via Motion Attention
• CAFE-GAN: Arbitrary Face Attribute Editing with Complementary Attention Feature
• Supervised Edge Attention Network for Accurate Image Instance Segmentation
• SPAN: Spatial Pyramid Attention Network for Image Manipulation Detection
• Forecasting Human-Object Interaction: Joint Prediction of Motor Attention and Actions in First Person Video
• An Attention-driven Two-stage Clustering Method for Unsupervised Person Re-Identification
• Spatiotemporal Attention Cell Search for Video Classification
• Learning Trailer Moments in Full-Length Movies with Co-Contrastive Attention
• Deep Surface Normal Estimation on the 2-Sphere with Confidence Guided Semantic Attention
• Attention-Based Query Expansion Learning
• Attention-Driven Dynamic Graph Convolutional Network for Multi-Label Image Recognition
• DA4AD: End-to-end Deep Attention Aware Features Aided Visual Localization for Autonomous Driving
• Few-shot Action Recognition via Permutation-invariant Attention
• Orientation-aware Vehicle Re-identification with Semantics-guided Part Attention Network
• Single Image Super-Resolution via a Holistic Attention Network
• SOLAR: Second-Order Loss and Attention for Image Retrieval
• GATCluster: Self-Supervised Gaussian-Attention Network for Image Clustering
• Cross-Attention in Coupled Unmixing Nets for Unsupervised Hyperspectral Super-Resolution
• Monocular Expressive Body Regression through Body-Driven Attention
• Deep Reinforced Attention Learning for Quality-Aware Visual Recognition
• Character Region Attention For Text Spotting
• Take an Emotion Walk: Perceiving Emotions from Gaits Using Hierarchical Attention Pooling and Affective Mapping
• End-to-End Low Cost Compressive Spectral Imaging with Spatial-Spectral Self-Attention
• Spatial Attention Pyramid Network for Unsupervised Domain Adaptation [paper]
• Semantic Line Detection Using Mirror Attention and Comparative Ranking and Matching
• Look here! A parametric learning based approach to redirect visual attention
• Guiding Monocular Depth Estimation Using Depth Attention-Volume
• Suppressing Mislabeled Data via Grouping and Self-Attention
• Weight Excitation: Built-in Attention Mechanisms in Convolutional Neural Networks
• Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation
• Few-Shot Semantic Segmentation with Democratic Attention Networks
• Empowering Relational Network by Self-Attention Augmented Conditional Random Fields for Group Activity Recognition
• AiR: Attention with Reasoning Capability
• Assembling Modality Representations via Attention Connections
• Box2Seg: Attention Weighted Loss and Discriminative Feature Learning for Weakly Supervised Segmentation
• Attention Guided Anomaly Localization in Images
• ReAD: Reciprocal Attention Discriminator for Image-to-Video Re-Identification
• Unsupervised Deep Metric Learning with Transformed Attention Consistency and Contrastive Clustering Loss
• The Devil is in the Details: Self-Supervised Attention for Vehicle Re-Identification
• Efficient Attention Mechanism for Visual Dialog that can Handle All the Interactions between Multiple Inputs
•  

[Super-Resolution]

• Component Divide-and-Conquer for Real-World Image Super-Resolution
• Face Super-Resolution Guided by 3D Facial Priors
• Learning with Privileged Information for Efficient Image Super-Resolution
• Fast Adaptation to Super-Resolution Networks via Meta-Learning
• PlugNet: Degradation Aware Scene Text Recognition Supervised by a Pluggable Super-Resolution Unit
• Towards Content-independent Multi-Reference Super-Resolution: Adaptive Pattern Matching and Feature Aggregation
• Single Image Super-Resolution via a Holistic Attention Network
• SRFlow: Learning the Super-Resolution Space with Normalizing Flow
• LatticeNet: Towards Lightweight Image Super-resolution with Lattice Block
• PAMS: Quantized Super-Resolution via Parameterized Max Scale
• Video Super-Resolution with Recurrent Structure-Detail Network
• Cross-Attention in Coupled Unmixing Nets for Unsupervised Hyperspectral Super-Resolution
• Journey Towards Tiny Perceptual Super-Resolution
• Across Scales & Across Dimensions: Temporal Super-Resolution using Deep Internal Learning
• Spatial-Angular Interaction for Light Field Image Super-Resolution
• Mining self-similarity: Label super-resolution with epitomic representations
• Stochastic Frequency Masking to Improve Super-Resolution and Denoising Networks
• MuCAN: Multi-Correspondence Aggregation Network for Video Super-Resolution
• VarSR: Variational Super-Resolution Network for Very Low Resolution Images
• Texture Hallucination for Large-Factor Painting Super-Resolution
• Zero-Shot Image Super-Resolution with Depth Guided Internal Degradation Learning
• Scene Text Image Super-Resolution in the Wild
• Component Divide-and-Conquer for Real-World Image Super-Resolution、

[Wavelet]

• Wavelet-Based Dual-Branch Neural Network for Image Demoireing
• Burst Denoising via Temporally Shifted Wavelet Transforms
• Multi-level Wavelet-based Generative Adversarial Network for Perceptual Quality Enhancement of Compressed Video
•  

[Bayesian]

• Unsupervised Bayesian Deep Learning for Image Reconstruction in Compressive Sensing
• AABO: Adaptive Anchor Box Optimization for Object Detection via Bayesian Sub-sampling
•  

[Semantic Segmentation]

• Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation (oral)
• Improving Semantic Segmentation via Decoupled Body and Edge Supervision
• SideInfNet: A Deep Neural Network for Semi-Automatic Semantic Segmentation with Side Information
• Content-Consistent Matching for Domain Adaptive Semantic Segmentation
• Label-Driven Reconstruction for Domain Adaptation in Semantic Segmentation
• Efficient Outdoor 3D Point Cloud Semantic Segmentation for Critical Road Objects and Distributed Contexts
• Classes Matter: A Fine-grained Adversarial Approach to Cross-domain Semantic Segmentation
• Virtual Multi-view Fusion for 3D Semantic Segmentation
• Bi-directional Cross-Modality Feature Propagation with Separation-and-Aggregation Gate for RGB-D Semantic Segmentation
• GMNet: Graph Matching Network for Large Scale Part Semantic Segmentation in the Wild
• Deep FusionNet for Point Cloud Semantic Segmentation
• Mining Cross-Image Semantics for Weakly Supervised Semantic Segmentation
• Few-Shot Semantic Segmentation with Democratic Attention Networks
• Prototype Mixture Models for Few-shot Semantic Segmentation
• Indirect Local Attacks for Context-aware Semantic Segmentation Networks
• Negative Pseudo Labeling using Class Proportion for Semantic Segmentation in Pathology
• Part-aware Prototype Network for Few-shot Semantic Segmentation
• Semi-supervised Semantic Segmentation via Strong-weak Dual-branch Network
• Self-Prediction for Joint Instance and Semantic Segmentation of Point Clouds
• Document Structure Extraction using Prior Based HighResolution Hierarchical Semantic Segmentation
• Learning from Scale-Invariant Examples for Domain Adaptation in Semantic Segmentation
• Contextual-Relation Consistent Domain Adaptation for Semantic Segmentation
• Object-Contextual Representations for Semantic Segmentation
• Learning to Predict Context-adaptive Convolution for Semantic Segmentation
• Domain Adaptive Semantic Segmentation Using Weak Labels
• EfficientFCN: Holistically-guided Decoding for Semantic Segmentation
• Splitting vs. Merging: Mining Object Regions with Discrepancy and Intersection Loss for Weakly Supervised Semantic Segmentation
• Weakly Supervised Semantic Segmentation with Boundary Exploration
• Increasing the Robustness of Semantic Segmentation Models with Painting-by-Numbers
• Class-wise Dynamic Graph Convolution for Semantic Segmentation
• JSENet: Joint Semantic Segmentation and Edge Detection Network for 3D Point Clouds
• Tensor Low-Rank Reconstruction for Semantic Segmentation
• Intra-class Compactness Distillation for Semantic Segmentation
• SNE-RoadSeg: Incorporating Surface Normal Information into Semantic Segmentation for Accurate Freespace Detection
• Employing Multi-Estimations for Weakly-Supervised Semantic Segmentation
•  
•  

[Distillation]

• StyleGAN2 Distillation for Feed-forward Image Manipulation
• Online Ensemble Model Compression using Knowledge Distillation
• Feature Normalized Knowledge Distillation for Image Classification
• Knowledge Distillation Meets Self-Supervision
• Discriminability Distillation in Group Representation Learning
• Optical Flow Distillation: Towards Efficient and Stable Video Style Transfer
• Prime-Aware Adaptive Distillation
• Robust Re-Identification by Multiple Views Knowledge Distillation
• HardGAN: A Haze-Aware Representation Distillation GAN for Single Image Dehazing
• DOPE: Distillation Of Part Experts for whole-body 3D pose estimation in the wild
• Weight Decay Scheduling and Knowledge Distillation for Active Learning
• Circumventing Outliers of AutoAugment with Knowledge Distillation
• Semantic Relation Preserving Knowledge Distillation for Image-to-Image Translation
• Domain Adaptation through Task Distillation
• Improving Face Recognition from Hard Samples via Distribution Distillation Loss
• Defocus Blur Detection via Depth Distillation
• Intra-class Compactness Distillation for Semantic Segmentation
• Differentiable Feature Aggregation Search for Knowledge Distillation
• Exclusivity-Consistency Regularized Knowledge Distillation for Face Recognition
• MetaDistiller: Network Self-boosting via Meta-learned Top-down Distillation
• Reversing the cycle: self-supervised deep stereo through enhanced monocular distillation
• Learning From Multiple Experts: Self-paced Knowledge Distillation for Long-tailed Classification
• Interpretable Foreground Object Search As Knowledge Distillation
• Improving Knowledge Distillation via Category Structure
• Local Correlation Consistency for Knowledge Distillation
• Knowledge Transfer via Dense Cross-layer Mutual-distillation
• Matching Guided Distillation
• AMLN: Adversarial-based Mutual Learning Network for Online Knowledge Distillation