Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Datasets/Kinetics

Kinetics

Kinetics Human Action Video Dataset

VideosCC BY 4.0Introduced 2017-05-19

The Kinetics dataset is a large-scale, high-quality dataset for human action recognition in videos. The dataset consists of around 500,000 video clips covering 600 human action classes with at least 600 video clips for each action class. Each video clip lasts around 10 seconds and is labeled with a single action class. The videos are collected from YouTube.

Source: Self-supervised Visual Feature Learning with Deep Neural Networks: A Survey

Benchmarks

Text-to-Video Generation/Accuracy Video/Top-1 Video Classification/Top-1 Visual Tracking/Average Jaccard Zero-Shot Action Recognition/Top-1 Accuracy Zero-Shot Action Recognition/Top-5 Accuracy

Related Benchmarks

Kinetics-100/Action Recognition/Accuracy Kinetics-100/Activity Recognition/Accuracy Kinetics-400/3D Action Recognition/Actions Top-1 (S1)Kinetics-400/Action Detection/Actions Top-1 (S1)Kinetics-400/Action Localization/Actions Top-1 (S1)Kinetics-400/Action Recognition/Actions Top-1 (S1)Kinetics-400/Action Recognition/Top-1 Accuracy Kinetics-400/Action Recognition/Top-1 accuracy %Kinetics-400/Action Recognition/Top-5 Accuracy Kinetics-400/Action Recognition/Top-5 Accuracy %Kinetics-400/Action Recognition In Videos/Top-1 Accuracy Kinetics-400/Action Recognition In Videos/Top-5 Accuracy Kinetics-400/Activity Recognition/Actions Top-1 (S1)Kinetics-400/Activity Recognition/Top-1 Accuracy Kinetics-400/Activity Recognition/Top-1 accuracy %Kinetics-400/Activity Recognition/Top-5 Accuracy Kinetics-400/Activity Recognition/Top-5 Accuracy %Kinetics-400/Boundary Detection/Pairwise F1 Kinetics-400/Boundary Detection/Precision Kinetics-400/Boundary Detection/Recall Kinetics-400/Event Segmentation/F1 Kinetics-400/Temporal Action Localization/Actions Top-1 (S1)Kinetics-400/Video/Acc@1 Kinetics-400/Video/Acc@5 Kinetics-400/Video/Actions Top-1 (S1)Kinetics-400/Video/Clip acc@1 Kinetics-400/Video/Clip acc@5 Kinetics-400/Video/FLOPs (G) x views Kinetics-400/Video/Parameters (M)Kinetics-400/Zero-Shot Learning/Actions Top-1 (S1)Kinetics-600/Action Recognition/Top-1 Accuracy Kinetics-600/Action Recognition/Top-5 Accuracy Kinetics-600/Action Recognition In Videos/Top-1 Accuracy Kinetics-600/Action Recognition In Videos/Top-5 Accuracy Kinetics-600/Activity Recognition/Top-1 Accuracy Kinetics-600/Activity Recognition/Top-5 Accuracy Kinetics-600/Video/GFLOPs Kinetics-600/Video/Top-1 Accuracy Kinetics-600/Video/Top-5 Accuracy Kinetics-600 12 frames, 128x128/Video/FID Kinetics-600 12 frames, 128x128/Video Generation/FID Kinetics-600 12 frames, 64x64/Video/Cond Kinetics-600 12 frames, 64x64/Video/FVD Kinetics-600 12 frames, 64x64/Video/IS Kinetics-600 12 frames, 64x64/Video/Pred Kinetics-600 12 frames, 64x64/Video Generation/FVD Kinetics-600 12 frames, 64x64/Video Prediction/Cond Kinetics-600 12 frames, 64x64/Video Prediction/FVD Kinetics-600 12 frames, 64x64/Video Prediction/IS Kinetics-600 12 frames, 64x64/Video Prediction/Pred Kinetics-600 48 frames, 64x64/Video/FID Kinetics-600 48 frames, 64x64/Video/Inception Score Kinetics-600 48 frames, 64x64/Video Generation/FID Kinetics-600 48 frames, 64x64/Video Generation/Inception Score Kinetics-700/Image Clustering/Accuracy Kinetics-700/Video/FID Kinetics-700/Video/FVD Kinetics-700/Video/Top-1 Accuracy Kinetics-700/Video/Top-5 Accuracy Kinetics-700/Video Generation/FID Kinetics-700/Video Generation/FVD Kinetics-700-2020/Video/Top 1 Accuracy Kinetics-GEB+/10-shot image generation/mAP Kinetics-GEB+/10-shot image generation/text-to-video R@1 Kinetics-GEB+/10-shot image generation/text-to-video R@10 Kinetics-GEB+/10-shot image generation/text-to-video R@5 Kinetics-GEB+/10-shot image generation/text-to-video R@50 Kinetics-GEB+/Text to Video Retrieval/mAP Kinetics-GEB+/Text to Video Retrieval/text-to-video R@1 Kinetics-GEB+/Text to Video Retrieval/text-to-video R@10 Kinetics-GEB+/Text to Video Retrieval/text-to-video R@5 Kinetics-GEB+/Text to Video Retrieval/text-to-video R@50 Kinetics-GEB+/Video/F1@0.1s Kinetics-GEB+/Video/F1@0.2s Kinetics-GEB+/Video/F1@0.5s Kinetics-GEB+/Video/F1@1.0s Kinetics-GEB+/Video/F1@1.5s Kinetics-GEB+/Video/F1@2.0s Kinetics-GEB+/Video/F1@2.5s Kinetics-GEB+/Video/F1@3.0s Kinetics-GEB+/Video/F1@Avg Kinetics-GEB+/Video Captioning/CIDEr Kinetics-GEB+/Video Captioning/ROUGE-L Kinetics-GEB+/Video Captioning/SPICE Kinetics-GEB+/Video Grounding/F1@0.1s Kinetics-GEB+/Video Grounding/F1@0.2s Kinetics-GEB+/Video Grounding/F1@0.5s Kinetics-GEB+/Video Grounding/F1@1.0s Kinetics-GEB+/Video Grounding/F1@1.5s Kinetics-GEB+/Video Grounding/F1@2.0s Kinetics-GEB+/Video Grounding/F1@2.5s Kinetics-GEB+/Video Grounding/F1@3.0s Kinetics-GEB+/Video Grounding/F1@Avg Kinetics-GEB+/Video Retrieval/F1@0.1s Kinetics-GEB+/Video Retrieval/F1@0.2s Kinetics-GEB+/Video Retrieval/F1@0.5s Kinetics-GEB+/Video Retrieval/F1@1.0s Kinetics-GEB+/Video Retrieval/F1@1.5s Kinetics-GEB+/Video Retrieval/F1@2.0s Kinetics-GEB+/Video Retrieval/F1@2.5s Kinetics-GEB+/Video Retrieval/F1@3.0s Kinetics-GEB+/Video Retrieval/F1@Avg Kinetics-GEBD/Event Segmentation/F1 @ RelDis. 0.05 Kinetics-Skeleton dataset/3D Action Recognition/Accuracy Kinetics-Skeleton dataset/3D Action Recognition/GFLOPS per prediction Kinetics-Skeleton dataset/Action Detection/Accuracy Kinetics-Skeleton dataset/Action Detection/GFLOPS per prediction Kinetics-Skeleton dataset/Action Localization/Accuracy Kinetics-Skeleton dataset/Action Localization/GFLOPS per prediction Kinetics-Skeleton dataset/Action Recognition/Accuracy Kinetics-Skeleton dataset/Action Recognition/GFLOPS per prediction Kinetics-Skeleton dataset/Activity Recognition/Accuracy Kinetics-Skeleton dataset/Activity Recognition/GFLOPS per prediction Kinetics-Skeleton dataset/Temporal Action Localization/Accuracy Kinetics-Skeleton dataset/Temporal Action Localization/GFLOPS per prediction Kinetics-Skeleton dataset/Video/Accuracy Kinetics-Skeleton dataset/Video/GFLOPS per prediction Kinetics-Skeleton dataset/Zero-Shot Learning/Accuracy Kinetics-Skeleton dataset/Zero-Shot Learning/GFLOPS per prediction Kinetics-Sounds/Video/Top 1 Accuracy Kinetics-Sounds/Video/Top 5 Accuracy

Statistics

Papers: 1,341
Benchmarks: 6

Links

Tasks

Action Classification Action Recognition Action Recognition In Videos Boundary Captioning Boundary Detection Boundary Grounding Event Segmentation Few Shot Action Recognition Generic Event Boundary Detection Image Clustering Long-tail Learning Self-Supervised Action Recognition Self-Supervised Action Recognition Linear Semantic Object Interaction Classification Skeleton Based Action Recognition Spatio-Temporal Action Localization Temporal Action Localization Text to Video Retrieval Text-to-Video Generation Video Video Captioning Video Classification Video Generation Video Grounding Video Prediction Video Recognition Video Retrieval Video Understanding Visual Tracking Zero-Shot Action Recognition imbalanced classification