Benchmark Almanac

Jester gesture recognition

RFEEN, 20 Crops 97.1% tops its leaderboard
Not in model cards

Models classify hand gestures in short videos. The task distinguishes movements with similar appearance but different directions or timing.

Made by
TwentyBN
Released
October 2019
Kind
Video understanding
Level
Everyday
Scored
Checked automatically
Answers
Prediction
Tasks
All public
Measures
Capability
Language
English
In and out
Video → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
In views
Vision
33
Entries in its paper, 0 models we track

RFEEN, 20 Crops tops the leaderboard with 97.1%

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.

  1. 1 RFEEN, 20 Crops 97.1%
  2. 2 Ford’s Gesture Recognition System 96.8%
  3. 3 DRX3D 96.6%
  4. 4 TSN two-stream aggregated with conv 96.4%
  5. 5 Spatiotemporal Two Streams network 96.3%
  6. 6 3D CNN Architecture 96.2%
  7. 7 Motion Feature Network (MFNet) 96.2%
  8. 8 RNP 96%
  9. 9 SSNet RGB resnet 95.8%
  10. 10 TVB 95.7%
  11. 11 Temporal Pyramid Relation Network for Video-Based Gesture Recognition 95.3%
  12. 12 DIN 95.3%
  13. 13 TRN - 8 segments 95%
  14. 14 3D CNN - Multi time scale evaluation 94.8%
  15. 15 8frames rgb 94.8%
  16. 16 TRN (CVPR’18 submission) 94.8%
  17. 17 TRN + BNInception 94.5%
  18. 18 slowfast res50 94.5%
  19. 19 3D CNN for transfer learning 94.3%
  20. 20 Besnet 94.2%
  21. 21 3D GesNet 94%
  22. 22 3D-GesNet(only rgb) 94%
  23. 23 ECO 93.8%
  24. 24 TRN-E 93.6%
  25. 25 One Stream Modified-I3D 93.4%
  26. 26 Modified C3D 89.3%
  27. 27 CNN+LSTM 86.3%
  28. 28 3D ResNet 101 86%
  29. 29 VideoLSTM 85.9%
  30. 30 3D convolutional neural network 85.5%
  31. 31 ConvLSTM 82.8%
  32. 32 3d+resnet18 81.5%
  33. 33 3D ResNet 68.1%

Source: The Jester gesture recognition paper

#EntryScoreSettingDateSource
1 RFEEN, 20 Crops 97.1% Official leaderboard
2 Ford’s Gesture Recognition System 96.8% Official leaderboard
3 DRX3D 96.6% Official leaderboard
4 TSN two-stream aggregated with conv 96.4% Official leaderboard
5 Spatiotemporal Two Streams network 96.3% Official leaderboard
6 3D CNN Architecture 96.2% Official leaderboard
7 Motion Feature Network (MFNet) 96.2% Official leaderboard
8 RNP 96% Official leaderboard
9 SSNet RGB resnet 95.8% Official leaderboard
10 TVB 95.7% Official leaderboard
11 Temporal Pyramid Relation Network for Video-Based Gesture Recognition 95.3% Official leaderboard
12 DIN 95.3% Official leaderboard
13 TRN - 8 segments 95% Official leaderboard
14 3D CNN - Multi time scale evaluation 94.8% Official leaderboard
15 8frames rgb 94.8% Official leaderboard
16 TRN (CVPR’18 submission) 94.8% Official leaderboard
17 TRN + BNInception 94.5% Official leaderboard
18 slowfast res50 94.5% Official leaderboard
19 3D CNN for transfer learning 94.3% Official leaderboard
20 Besnet 94.2% Official leaderboard
21 3D GesNet 94% Official leaderboard
22 3D-GesNet(only rgb) 94% Official leaderboard
23 ECO 93.8% Official leaderboard
24 TRN-E 93.6% Official leaderboard
25 One Stream Modified-I3D 93.4% Official leaderboard
26 Modified C3D 89.3% Official leaderboard
27 CNN+LSTM 86.3% Official leaderboard
28 3D ResNet 101 86% Official leaderboard
29 VideoLSTM 85.9% Official leaderboard
30 3D convolutional neural network 85.5% Official leaderboard
31 ConvLSTM 82.8% Official leaderboard
32 3d+resnet18 81.5% Official leaderboard
33 3D ResNet 68.1% Official leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Audio and video

All 695 in Audio and video
  • Video-MMMU In use 300 expert-level lecture videos with questions testing perception, comprehension and adaptation of knowledge. Accuracy. Qwen 3.8 Max 88.7%
  • LVBench In use Hours-long video understanding with multiple-choice questions on reasoning, grounding and summarization. Gemini 4 Argon 91.7%
  • MMVU In use 3,000 expert-annotated questions on videos across 27 disciplines. Multiple-choice and open-ended accuracy. Qwen 3.8 Max 82.4%
  • MLVU In use Multi-task long video understanding (minutes to hours) with multiple-choice and generation tasks. Accuracy. Qwen 3.8 Max 90.8%
  • FLEURS In use Models transcribe read speech in 102 languages built from FLoRes-101 sentences. Model cards usually report word error rate, so lower scores are better. Phi-4-multimodal 4.00%
  • VoiceBench In use Spoken instructions with speaker, environment and content variation. Voice assistant responses scored by GPT judge or rules. Qwen3.5-Omni-Plus 93.1%

Data audit

All 8 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
  • Pass Independent runs Scores from the benchmark’s paper (33 entries).
  • Pass Models scored 33 entries on its leaderboard, none a model we track.
  • Note Newest score No scores yet.
  • Note Frontier models No models scored yet.
  • Note Headroom No scores yet.
Leaderboard import 3 pass
  • Pass Importer We read the board on openaccess.thecvf.com: 33 entries at the last read, none a model we track.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check