AI Paper Atlas

Research map / Benchmarks / AVA v2.2

AVA v2.2 leaderboard

19 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on AVA v2.2. Best mAP: 45.1 (LART-Hiera, CVPR 2023).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

AVA v2.2 · mAP

19 papers, each with its best reported number for its own method; higher is better.

#mAP ↑MethodPaperTopic
145.1LART-Hiera
Hiera
On the Benefits of 3D Pose and Tracking for Human Action Recognition
CVPR 2023 · Table: Comparison with state-o
Human pose estimation
242.5STAR/L
ViViT-L
End-to-End Spatio-Temporal Action Localisation with Video Transformers
CVPR 2024 · Table 9
Action recognition
341.1MVD-H
ViT-Huge
Masked Video Distillation: Rethinking Masked Feature Modeling for Self-Supervised Video Representation Learning
CVPR 2023 · AVA v2.2 comparison table
Video segmentation and tracking
439.8UMT-L (K710)Unmasked Teacher: Towards Training-Efficient Video Foundation Models
ICCV 2023 · Table 10
Video-language understanding
539.7EVAD rho=0.7
ViT-L (K700)
Efficient Video Action Detection with Token Dropout and Context Refinement
ICCV 2023 · Table (SOTA comparison)
Action recognition
639.5STMixer (ViT-L, K700)
ViT-L
STMixer: A One-Stage Sparse Action Detector
CVPR 2023 · Table: Comparison with state-o
Action recognition
739.5VideoMAE ViT-H (K400, with labels)
ViT-H
VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training
NeurIPS 2022 · Table AVA
Video segmentation and tracking
835.4MeMViT-24, 32×3, 312²
MeMViT-24, 32×3
MeMViT: Memory-Augmented Multiscale Vision Transformer for Efficient Long-Term Video Recognition
CVPR 2022 · Table (AVA comparison)
Action recognition
934.42AIA R101*, 3 ens.
K700
Asynchronous Interaction Aggregation for Action Detection
ECCV 2020 · Table 2b
Action recognition
1034.3SlowFast++ ensembleSlowFast Networks for Video Recognition
ICCV 2019 · Table 7
Action recognition
9 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.