欧洲计算机视觉国际会议ECCR20最新论文整理分享_. online action detectionin european conference on

    ECCV的全称是European Conference on Computer Vision(欧洲计算机视觉国际会议) ，两年一次，是计算机视觉三大会议（另外两个是ICCV和CVPR）之一。每次会议在全球范围录用论文300篇左右，主要的录用论文都来自美国、欧洲等顶尖实验室及研究所，中国大陆的论文数量一般在10-20篇之间。

    本文整理了2020年ECCV所有Accept的论文，论文完整版源地址：https://eccv2020.eu/accepted-papers/

 

精选论文列表

    Quaternion Equivariant Capsule Networks for 3D Point Clouds

    

    DeepFit: 3D Surface Fitting by Neural Network Weighted Least Squares

    

    MoSaNAS: Multi-Objective Surrogate-Assisted Neural Architecture Search

    

    Describing Textures using Natural Language

    

    Empowering Relational Network by Self-Attention Augmented Conditional Random Fields for Group Activity Recognition

    

    AiR: Attention with Reasoning Capability

    

    Self6D: Self-Supervised Monocular 6D Object Pose Estimation

    

    Invertible Image Rescaling

    

    Synthesize then Compare: Detecting Failures and Anomalies for Semantic Segmentation

    

    House-GAN: Relational Generative Adversarial Networks for Graph-constrained House Layout Generation

    

    Crowdsampling the Plenoptic Function

    

    End-to-End Estimation of Multi-Person 3D Poses from Multiple Cameras

    

    End-to-End Object Detection with Transformers

    

    DeepSFM: Structure From Motion Via Deep Bundle Adjustment

    

    Ladybird: Deep Implicit Field Based 3D Reconstruction with Sampling and Symmetry

    

    Segment as Points for Efficient Online Multi-Object Tracking and Segmentation

    

    Conditional Convolutions for Instance Segmentation

    

    MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution

    

    Fashionpedia: Ontology, Segmentation, and an Attribute Localization Dataset

    

    Privacy Preserving Structure-from-Motion

    

    Rewriting a Deep Generative Model

    

    Compare and Reweight: Distinctive Image Captioning Using Similar Images Sets

    

    Long-term Human Motion Prediction with Scene Context

    

    NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis

    

    ReferIt3D: Neural Listeners for Fine-Grained 3D Object Identification in Real-World Scenes

    

    MatryODShka: Real-time 6DoF Video View Synthesis using Multi-Sphere Images

    

    Learning and aggregating deep local descriptors for instance-level recognition

    

    A Consistently Fast and Globally Optimal Solution to the Perspective-n-Point Problem

    

    Learn to Recover Visible Color for Video Surveillance in a Day

    

    Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction from Single-view Images

    

    Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation

    

    BorderDet: Border Feature for Dense Object Detection

    

    Regularization with Latent Space Virtual Adversarial Training

    

    Du$^2$Net: Learning Depth Estimation from Dual-Cameras and Dual-Pixels

    

    Model-Agnostic Boundary-Adversarial Sampling for Test-Time Generalization in Few-Shot learning

    

    Targeted Attack for Deep Hashing based Retrieval

    

    Gradient Centralization: A New Optimization Technique for Deep Neural Networks

    

    Content-Aware Unsupervised Deep Homography Estimation

    

    Multi-View Optimization of Local Feature Geometry

    

    Efï¬cient Model Fitting by Combining Lifted Optimization with Phong Surface Models

    

    Forecasting Human-Object Interaction: Joint Prediction of Motor Attention and Actions in First Person Video

    

    Learning Stereo from Single Images

    

    Prototype Rectification for Few-Shot Learning

    

    Learning Feature Descriptors using Camera Pose Supervision

    

    Semantic Flow for Fast and Accurate Scene Parsing

    

    Appearance Consensus Driven Self-Supervised Human Mesh Recovery

    

    Diffraction Line Imaging

    

    Aligning and Projecting Images to Class-conditional Generative Networks

    

    Suppress and Balance: A Simple Gated Network for Salient Object Detection

    

    Visual Memorability for Robotic Interestingness Prediction via Unsupervised Online Learning

    

    Post-Training Piecewise Linear Quantization for Deep Neural Networks

    

    Joint Disentangling and Adaptation for Cross-Domain Person Re-Identification

    

    In-Home Daily-Life Captioning Using Radio Signals

    

    Self-Challenging Improves Cross-Domain Generalization

    

    A Competence-aware Curriculum for Visual Concepts Learning via Question Answering

    

    Multi-task Learning Increases Adversarial Robustness

    

    S2DNAS: Transforming Static CNN Model for Dynamic Inference via Neural Architecture Search

    

    Improving Deep Video Compression by Resolution-adaptive Flow Coding

    

    Motion Capture from Internet Videos

    

    Appearance-Preserving 3D Convolution for Video-based Person Re-identification

    

    Solving the Blind Perspective-n-Point Problem End-To-End With Robust Differentiable Geometric Optimization

    

    Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulation

    

    Deep Spatial-angular Regularization for Compressive Light Field Reconstruction over Coded Apertures

    

    Video-based Remote Physiological Measurement via Cross-verified Feature Disentangling

    

    Combining Implicit Function Learning and Parametric Models for 3D Human Reconstruction

    

    Orientation-aware Vehicle Re-identification with Semantics-guided Part Attention Network

    

    Mining Cross-Image Semantics for Weakly Supervised Semantic Segmentation

    

    Coherent full scene 3D reconstruction from a single RGB image

    

    Layer-wise Conditioning Analysis in Exploring the Learning Dynamics of DNNs

    

    RAFT: Recurrent All-Pairs Field Transforms for Optical Flow

    

    Domain-invariant Stereo Matching Networks

    

    DeepHandMesh: Weakly-supervised Deep Encoder-Decoder Framework for High-fidelity Hand Mesh Modeling from a Single RGB Image

    

    Content Adaptive and Error Propagation Aware Deep Video Compression

    

    Towards Streaming Image Understanding

    

    Towards Automated Testing and Robustification by Semantic Adversarial Data Generation

    

    Adversarial Generative Grammars for Human Activity Prediction

    

    Greedy Sampler and Dumb Learner: A Surprisingly Effective Approach for Continual Learning

    

    Learning Lane Graph Representations for Motion Forecasting

    

    What Matters in Unsupervised Optical Flow

    

    Synthesis and Completion of Facades from Satellite Imagery

    

    Mapillary Planet-Scale Depth Dataset

    

    V2VNet: Vehicle-to-Vehicle Communication for Joint Perception and Prediction

    

    Training Interpretable Convolutional Neural Networks by Differentiating Class-specific Filters

    

    EagleEye: Fast Sub-net Evaluation for Efficient Neural Network Pruning

    

    Intrinsic Point Cloud Interpolation via Dual Latent Space Navigation

    

    Cross-Domain Cascaded Deep Translation

    

    "Look Ma, no landmarks!" - Unsupervised, model-based dense face alignment

    

    Online Invariance Selection for Local Feature Descriptors

    

    Rethinking image inpainting via a mutual encoder-decoder with feature equalization

    

    TextCaps: a Dataset for Image Captioning with Reading Comprehension

    

    It is not the Journey but the Destination: Endpoint Conditioned Trajectory Prediction

    

    Learning What to Learn for Video Object Segmentation

    

    SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size Sensitive 3D Clothing

    

    LIMP: Learning Latent Shape Representations with Metric Preservation Priors

    

    Unsupervised Sketch-to-Photo Synthesis

    

    A simple way to make neural networks robust against diverse image corruptions

    

    SoftpoolNet: Shape Descriptor for Point Cloud Completion and Classification

    

    Hierarchical Face Aging through Disentangled Latent Characteristics

    

    Hybrid Models for Open Set Recognition

    

    TopoGAN: A Topology-Aware Generative Adversarial Network

    

    Learning to Localize Actions from Moments

    

    ForkGAN: Seeing into the Rainy Night

    

    TCGM: An Information-Theoretic Framework for Semi-Supervised Multi-Modality Learning

    

    ExchNet: A Unified Hashing Network for Large-Scale Fine-Grained Image Retrieval

    

    A Simple and Versatile Framework for Image-to-Image Translation

    

    ProxyBNN: Learning Binarized Neural Networks via Proxy Matrices

    

    Fair Attribute Classification through Latent Space De-biasing

    

    HMOR: Hierarchical Multi-person Ordinal Relations for Monocular Multi-Person 3D Pose Estimation

    

    Mask2CAD: 3D Shape Prediction by Learning to Segment and Retrieve

    

    A Unified Framework of Surrogate Loss by Refactorization and Interpolation

    

    Deep Reflectance Volumes: Relightable Reconstructions from Multi-View Photometric Images

    

    Memory-augmented Dense Predictive Coding for Video Representation Learning

    

    PointMixup: Augmentation for Point Clouds

    

    Identity-Guided Human Semantic Parsing Learning for Person Re-Identification

    

    Learning Gradient Fields for Shape Generation

    

    Few-Shot Unsupervised Image Translation with a Content Conditioned Style Encoder

    

    Corner Proposal Network for Anchor-free, Two-stage Object Detection

    

    PhraseClick: Toward Achieving Flexible Interactive Segmentation by Phrase and Click

    

    Unified Multisensory Perception: Weakly-Supervised Audio-Visual Video Parsing

    

    Learning Delicate Local Representations for Multi-Person Pose Estimation

    

    Learning to plan with uncertain topological maps

    

    Neural Design Network: Graphic Layout Generation with Constraints

    

    Learning Open Set Network with Discriminative Reciprocal Points

    

    Convolutional Occupancy Networks

    

    Multi-person 3D Pose Estimation in Crowded Scenes Based on Multi-View Geometry

    

    A General Toolbox for Understanding Errors in Object Detection

    

    PointContrast: Unsupervised Pretraining for 3D Point Cloud Understanding

    

    DSA: More Efficient Budgeted Pruning via Differentiable Sparsity Allocation

    

    Circumventing Outliers of AutoAugment with Knowledge Distillation

    

    S2DNet: Learning accurate correspondences for sparse-to-dense feature matching

    

    RTM3D: Real-time Monocular 3D Detection from Object Keypoints for Autonomous Driving

    

    Video Object Segmentation with Graph Memory Network

    

    Rethinking Bottleneck Structure for Efficient Mobile Network Design

    

    Side-Tuning: A Baseline for Network Adaptation via Additive Side Networks

    

    Towards Part-aware Monocular 3D Human Pose Estimation: An Architecture Search Approach

    

    A Tool for Measuring and Mitigating Bias in Visual Datasets

    

    Contrastive Learning for Weakly Supervised Phrase Grounding

    

    Collaborative Learning of Gesture Recognition and 3D Hand Pose Estimation with Multi-Order Feature Analysis

    

    Studying the Transferability of Adversarial Attacks on Object Detectors

    

    TuiGAN: Learning Versatile Image-to-Image Translation with Two Unpaired Images

    

    Semi-Siamese Training for Shallow Face Learning

    

    GAN Slimming: All-in-One Unified GAN Compression

    

    Human Interaction Learning on 3D Skeleton Point Clouds for Video Violence Recognition

    

    Binarized Neural Network for Single Image Super Resolution

    

    Axial-DeepLab: Stand-Alone Axial-Attention for Panoptic Segmentation

    

    Adaptive Computationally Efficient Network for Monocular 3D Hand Pose Estimation

    

    Chained-Tracker: Chaining Paired Attentive Regression Results for End-to-End Joint Multiple-Object Detection and Tracking

    

    Distribution-Balanced Loss for Multi-Label Classification in Long-Tailed Datasets

    

    Hamiltonian Dynamics for Real-World Shape Interpolation

    

    Learning to Scale Multilingual Representations for Vision-Language Tasks