README.md

Publications in ICCV 2021

Scene Graphs

• Graph-to-3D: End-to-End Generation and Manipulation of 3D Scenes Using Scene Graphs
• DeepPanoContext: Panoramic 3D Scene Understanding with Holistic Scene Context Graph and Relation-based Optimization
• Visual Scene Graphs for Audio Source Separation
• OSCAR-Net: Object-centric Scene Graph Attention for Image Attribution
• In Defense of Scene Graphs for Image Captioning
• Holistic Pose Graph: Modeling Geometric Structure Among Objects in a Scene Using Graph Inference for 3D Object Prediction

Scene Graph Generation

• Generative Compositional Augmentations for Scene Graph Prediction
• Unconditional Scene Graph Generation
• A Simple Baseline for Weakly-Supervised Scene Graph Generation

Human Interaction Understanding

• Exploiting Scene Graphs for Human Interaction Detection
• Consistency-Aware Graph Network for Human Interaction Understanding
• Spatially Conditioned Graphs for Detecting Human-Object Interactions
• H2O: Two Hands Manipulating Objects for First Person Interaction Recognition

Object Detection

• ODAM: Object Detection, Association, and Mapping using Posed RGB Video
• GraphFPN: Graph Feature Pyramid Network for Object Detection
• Deep Structured Instance Graph for Distilling Object Detectors
• Dual Bipartite Graph Learning: A General Approach for Domain Adaptive Object Detection
• CaT: Weakly Supervised Object Detection With Category Transfer
• Light Field Saliency Detection With Dual Local Graph Learning and Reciprocative Guidance

3D Object Detection

• Multi-Echo LiDAR for 3D Object Detection
• Pyramid R-CNN: Towards Better Performance and Adaptability for 3D Object Detection

Few-Shot Learning

• Query Adaptive Few-Shot Object Detection With Heterogeneous Graph Convolutional Networks
• Hierarchical Graph Attention Network for Few-Shot Visual-Semantic Learning

Semi-Supervised Learning

• Semi-Supervised Active Learning for Semi-Supervised Models: Exploit Adversarial Examples With Graph-Based Virtual Labels
• Graph-BAS3Net: Boundary-Aware Semi-Supervised Segmentation Network With Bilateral Graph Convolution

Image Recognition

• Transformer-Based Dual Relation Graph for Multi-Label Image Recognition
• Graph-Based Asynchronous Event Processing for Rapid Object Recognition

Spatial-Temporal Structures

• Spatio-Temporal Dynamic Inference Network for Group Activity Recognition
• End-to-End Video Instance Segmentation via Spatial-Temporal Graph Neural Networks
• Unifying Nonlocal Blocks for Neural Networks
• Spatial-Temporal Consistency Network for Low-Latency Trajectory Forecasting

Pose Estimation

• Graph-Based 3D Multi-Person Pose Estimation Using Multi-View Images
• PR-GCN: A Deep Graph Convolutional Network with Point Refinement for 6D Pose Estimation
• Learning Skeletal Graph Neural Networks for Hard 3D Pose Estimation
• Modulated Graph Convolutional Network for 3D Human Pose Estimation
• PoGO-Net: Pose Graph Optimization With Graph Neural Networks
• Space-Time-Separable Graph Convolutional Network for Pose Forecasting
• Mesh Graphormer

Skeleton Action Recognition

• Channel-wise Topology Refinement Graph Convolution for Skeleton-Based Action Recognition
• Skeleton Cloud Colorization for Unsupervised 3D Action Representation Learning

Trajectory Prediction

• Unlimited Neighborhood Interaction for Heterogeneous Trajectory Prediction
• Human Trajectory Prediction via Counterfactual Analysis
• TRiPOD: Human Trajectory and Pose Dynamics Forecasting in the Wild

Videos

• Adaptive Hierarchical Graph Reasoning with Semantic Coherence for Video-and-Language Inference
• HAIR: Hierarchical Visual-Semantic Relational Reasoning for Video Question Answering
• MAAS: Multi-Modal Assignation for Active Speaker Detection
• Unified Graph Structured Models for Video Understanding
• Video Self-Stitching Graph Network for Temporal Action Localization
• Multi-Modal Multi-Action Video Recognition
• Motion Guided Region Message Passing for Video Captioning
• Click To Move: Controlling Video Generation With Sparse Motion
• HighlightMe: Detecting Highlights From Human-Centric Videos

Matting

• Video Matting via Consistency-Regularized Graph Neural Networks
• Cascade Image Matting with Deformable Graph Refinement

Graph Matching

• Learning to Match Features with Seeded Graph Matching Network
• Hypergraph Neural Networks for Hypergraph Matching

Point Clouds

• Adaptive Graph Convolution for Point Cloud Analysis
• Towards Efficient Graph Convolutional Networks for Point Cloud Handling
• Perturbed Self-Distillation: Weakly Supervised Large-Scale Point Cloud Semantic Segmentation
• Distinctiveness Oriented Positional Equilibrium for Point Cloud Registration
• Persistent Homology Based Graph Convolution Network for Fine-Grained 3D Shape Segmentation
• Free-form Description Guided 3D Visual Graph Network for Object Grounding in Point Cloud

Navigation

• Hierarchical Object-to-Zone Graph for Object Navigation
• Visual Graph Memory With Unsupervised Representation for Visual Navigation

Image Clustering

• Learning Hierarchical Graph Neural Networks for Image Clustering
• Learn To Cluster Faces via Pairwise Classification

Image Restoration

• Dynamic Attentive Graph Learning for Image Restoration
• Cross-Patch Graph Convolutional Network for Image Denoising

Motion Prediction

• MSR-GCN: Multi-Scale Residual Graph Convolution Networks for Human Motion Prediction
• Motion Prediction Using Trajectory Cues

Person Re-identification

• Attack-Guided Perceptual Data Generation for Real-World Re-Identification
• Meta Pairwise Relationship Distillation for Unsupervised Person Re-Identification

Buildings

• Building-GAN: Graph-Conditioned Architectural Volumetric Design Generation
• BuildingNet: Learning To Label 3D Buildings

Vehicles

• Heterogeneous Relational Complement for Vehicle Re-identification
• Track Without Appearance: Learn Box and Tracklet Embedding With Local and Global Motion Patterns for Vehicle Tracking

Deep Metric Learning

• Deep Relational Metric Learning
• Learning Better Visual Data Similarities via New Grouplet Non-Euclidean Embedding

Domain Adaptation

• Towards Novel Target Discovery Through Open-Set Domain Adaptation
• Shape-Biased Domain Generalization via Shock Graph Embeddings

Architecture Search

• Not All Operations Contribute Equally: Hierarchical Operation-Adaptive Predictor for Neural Architecture Search
• Auto Graph Encoder-Decoder for Neural Network Pruning

Miscellaneous

• Distilling Holistic Knowledge with Graph Neural Networks
• RAIN: Reinforced Hybrid Attention Inference Network for Motion Forecasting
• SketchLattice: Latticed Representation for Sketch Manipulation
• Unsupervised Dense Deformation Embedding Network for Template-Free Shape Correspondence
• TGRNet: A Table Graph Reconstruction Network for Table Structure Recognition
• GP-S3Net: Graph-based Panoptic Sparse Semantic Segmentation Network
• Meta-Aggregator: Learning to Aggregate for 1-bit Graph Neural Networks
• RDI-Net: Relational Dynamic Inference Networks
• Adversarial Example Detection Using Latent Neighborhood Graph
• FloorPlanCAD: A Large-Scale CAD Drawing Dataset for Panoptic Symbol Spotting
• Adaptive Boundary Proposal Network for Arbitrary Shape Text Detection
• Enhanced Boundary Learning for Glass-Like Object Segmentation
• BlockPlanner: City Block Generation With Vectorized Graph Representation
• Wasserstein Coupled Graph Learning for Cross-Modal Retrieval
• Pose Invariant Topological Memory for Visual Navigation