NeurIPS 2023 accepted papers on Video

85 of the 3,584 papers accepted at NeurIPS 2023 (2.4%) are on Video, compared with 2.0% at NeurIPS 2022. They fall into 4 topics.

See NeurIPS 2023 in the interactive map

Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.

Topics and papers

Video segmentation and tracking · 41 papers

OpenSTL: A Comprehensive Benchmark of Spatio-Temporal Predictive Learning (32 citations)
HiNeRV: Video Compression with Hierarchical Encoding-based Neural Representation (22 citations)
Cinematic Mindscapes: High-quality Video Reconstruction from Brain Activity (20 citations)
MAViL: Masked Audio-Video Learners (17 citations)
BiMatting: Efficient Video Matting via Binarization (14 citations)
GPT-ST: Generative Pre-Training of Spatio-Temporal Graph Neural Networks (14 citations)
STARSS23: An Audio-Visual Dataset of Spatial Recordings of Real Scenes with Spatiotemporal Annotations of Sound Events (13 citations)
Video Prediction Models as Rewards for Reinforcement Learning (11 citations)
StressID: a Multimodal Dataset for Stress Identification (10 citations)
ChimpACT: A Longitudinal Dataset for Understanding Chimpanzee Behaviors (6 citations)
Diversifying Spatial-Temporal Perception for Video Domain Generalization (6 citations)
Mr. HiSum: A Large-scale Dataset for Video Highlight Detection and Summarization (6 citations)

and 29 more in the interactive map.

Video-language understanding · 20 papers

EgoSchema: A Diagnostic Benchmark for Very Long-form Video Language Understanding (45 citations)
Self-Chained Image-Language Model for Video Localization and Question Answering (30 citations)
MomentDiff: Generative Video Moment Retrieval from Random to Real (23 citations)
Perception Test: A Diagnostic Benchmark for Multimodal Video Models (17 citations)
SOC: Semantic-Assisted Object Cluster for Referring Video Object Segmentation (14 citations)
Paxion: Patching Action Knowledge in Video-Language Foundation Models (9 citations)
Learning Fine-grained View-Invariant Representations from Unpaired Ego-Exo Videos via Temporal Alignment (7 citations)
Auslan-Daily: Australian Sign Language Translation for Daily Communication and News (6 citations)
EgoDistill: Egocentric Head Motion Distillation for Efficient Video Understanding (6 citations)
Single-Stage Visual Query Localization in Egocentric Videos (6 citations)
EgoEnv: Human-centric environment representations from egocentric video (4 citations)
VidChapters-7M: Video Chapters at Scale (4 citations)

and 8 more in the interactive map.

Action recognition · 13 papers

Video-Mined Task Graphs for Keystep Recognition in Instructional Videos (10 citations)
Unsupervised Video Domain Adaptation for Action Recognition: A Disentanglement Perspective (9 citations)
HA-ViD: A Human Assembly Video Dataset for Comprehensive Assembly Knowledge Understanding (7 citations)
Language-based Action Concept Spaces Improve Video Self-Supervised Learning (7 citations)
EPIC Fields: Marrying 3D Geometry and Video Understanding (6 citations)
CAST: Cross-Attention in Space and Time for Video Action Recognition (5 citations)
Ego4D Goal-Step: Toward Hierarchical Understanding of Procedural Activities (4 citations)
NurViD: A Large Expert-Level Video Database for Nursing Procedure Activity Understanding (4 citations)
Activity Grammars for Temporal Action Segmentation (2 citations)
On Occlusions in Video Action Detection: Benchmark Datasets And Training Recipes (2 citations)
Revealing the unseen: Benchmarking video action recognition under occlusion (2 citations)
Learning Human Action Recognition Representations Without Real Humans (1 citation)

and 1 more in the interactive map.

NeurIPS 2023 accepted papers on Video

Topics and papers

Video segmentation and tracking · 41 papers

Video-language understanding · 20 papers

Action recognition · 13 papers

Video generation · 11 papers

More