SOTAVerified

Action Recognition

Action Recognition is a computer vision task that involves recognizing human actions in videos or images. The goal is to classify and categorize the actions being performed in the video or image into a predefined set of action classes.

In the video domain, it is an open question whether training an action classification network on a sufficiently large dataset, will give a similar boost in performance when applied to a different temporal task or dataset. The challenges of building video datasets has meant that most popular benchmarks for action recognition are small, having on the order of 10k videos.

Please note some benchmarks may be located in the Action Classification or Video Classification tasks, e.g. Kinetics-400.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 701–750 of 2759 papers

TitleStatusHype
Modality Distillation with Multiple Stream Networks for Action RecognitionCode0
MMG-Ego4D: Multimodal Generalization in Egocentric Action RecognitionCode0
Modeling the Relative Visual Tempo for Self-supervised Skeleton-based Action RecognitionCode0
Back to the Future: Cycle Encoding Prediction for Self-supervised Contrastive Video Representation LearningCode0
MMG-Ego4D: Multi-Modal Generalization in Egocentric Action RecognitionCode0
MaCLR: Motion-aware Contrastive Learning of Representations for VideosCode0
Mining YouTube - A dataset for learning fine-grained action concepts from webly supervised video dataCode0
3D CNNs on Distance Matrices for Human Action RecognitionCode0
Mission Balance: Generating Under-represented Class Samples using Video Diffusion ModelsCode0
MFAS: Multimodal Fusion Architecture SearchCode0
Adversarial Augmentation Training Makes Action Recognition Models More Robust to Realistic Video Distribution ShiftsCode0
MITFAS: Mutual Information based Temporal Feature Alignment and Sampling for Aerial Video Action RecognitionCode0
MOFO: MOtion FOcused Self-Supervision for Video UnderstandingCode0
Memory Attention Networks for Skeleton-based Action RecognitionCode0
MD-BERT: Action Recognition in Dark Videos via Dynamic Multi-Stream Fusion and Temporal ModelingCode0
Mask and Compress: Efficient Skeleton-based Action Recognition in Continual LearningCode0
Meta-path Analysis on Spatio-Temporal Graphs for Pedestrian Trajectory PredictionCode0
A Video-based End-to-end Pipeline for Non-nutritive Sucking Action Recognition and Segmentation in Young InfantsCode0
MARINE: A Computer Vision Model for Detecting Rare Predator-Prey Interactions in Animal VideosCode0
Advancing Compressed Video Action Recognition through Progressive Knowledge DistillationCode0
A Variational Time Series Feature Extractor for Action PredictionCode0
Mamba Fusion: Learning Actions Through QuestioningCode0
DMCL: Distillation Multiple Choice Learning for Multimodal Action RecognitionCode0
A^2-Nets: Double Attention NetworksCode0
MARS: Motion-Augmented RGB Stream for Action RecognitionCode0
Domain and View-point Agnostic Hand Action RecognitionCode0
MetaVD: A Meta Video Dataset for enhancing human action recognition datasetsCode0
A Variational Graph Autoencoder for Manipulation Action Recognition and PredictionCode0
LSTA: Long Short-Term Attention for Egocentric Action RecognitionCode0
LSTA-Net: Long short-term Spatio-Temporal Aggregation Network for Skeleton-based Action RecognitionCode0
Low-light Environment Neural SurveillanceCode0
Long-term Temporal Convolutions for Action RecognitionCode0
Long-Term Feature Banks for Detailed Video UnderstandingCode0
LSFB-CONT and LSFB-ISOL: Two New Datasets for Vision-Based Sign Language RecognitionCode0
Local Spherical Harmonics Improve Skeleton-Based Hand Action RecognitionCode0
LoCATe-GAT: Modeling Multi-Scale Local Context and Action Relationships for Zero-Shot Action RecognitionCode0
Lightweight Recurrent Cross-modal Encoder for Video Question AnsweringCode0
ADG-Pose: Automated Dataset Generation for Real-World Human Pose EstimationCode0
AutoGCN -- Towards Generic Human Activity Recognition with Neural Architecture SearchCode0
Making Third Person Techniques Recognize First-Person Actions in Egocentric VideosCode0
DVANet: Disentangling View and Action Features for Multi-View Action RecognitionCode0
Learn to cycle: Time-consistent feature discovery for action recognitionCode0
Learning Visual Actions Using Multiple Verb-Only LabelsCode0
Learning with privileged information via adversarial discriminative modality distillationCode0
Let's Dance: Learning From Online Dance VideosCode0
MLP-3D: A MLP-like 3D Architecture with Grouped Time MixingCode0
Learning Video Representations from Correspondence ProposalsCode0
Learning to Estimate Pose by Watching VideosCode0
MMTM: Multimodal Transfer Module for CNN FusionCode0
Discriminating Spatial and Temporal Relevance in Deep Taylor Decompositions for Explainable Activity RecognitionCode0
Show:102550
← PrevPage 15 of 56Next →
All datasetsSomething-Something V2UCF101HMDB-51Something-Something V1AVA v2.2EPIC-KITCHENS-100NTU RGB+DNTU RGB+D 120Diving-48ActivityNetAVA v2.1H2O (2 Hands and Objects)

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1MViTv2-B (IN-21K + Kinetics400 pretrain)Top-5 Accuracy93.4—Unverified
2RSANet-R50 (8+16 frames, ImageNet pretrained, 2 clips)Top-5 Accuracy91.1—Unverified
3MVD (Kinetics400 pretrain, ViT-H, 16 frame)Top-1 Accuracy77.3—Unverified
4InternVideoTop-1 Accuracy77.2—Unverified
5DejaVidTop-1 Accuracy77.2—Unverified
6InternVideo2-1BTop-1 Accuracy77.1—Unverified
7VideoMAE V2-gTop-1 Accuracy77—Unverified
8MVD (Kinetics400 pretrain, ViT-L, 16 frame)Top-1 Accuracy76.7—Unverified
9Hiera-L (no extra data)Top-1 Accuracy76.5—Unverified
10TubeViT-LTop-1 Accuracy76.1—Unverified
#ModelMetricClaimedVerifiedStatus
1FTP-UniFormerV2-L/143-fold Accuracy99.7—Unverified
2OmniVec23-fold Accuracy99.6—Unverified
3OmniVec3-fold Accuracy99.6—Unverified
4VideoMAE V2-g3-fold Accuracy99.6—Unverified
5BIKE3-fold Accuracy98.8—Unverified
6SMART3-fold Accuracy98.64—Unverified
7ZeroI2V ViT-L/143-fold Accuracy98.6—Unverified
8OmniSource (SlowOnly-8x8-R101-RGB + I3D-Flow)3-fold Accuracy98.6—Unverified
9PERF-Net (multi-distilled S3D)3-fold Accuracy98.6—Unverified
10Text4Vis3-fold Accuracy98.2—Unverified