Benchmark Almanac

FreiHAND

ExtPose 4.90 tops its leaderboard
Not in model cards

Models recover a hand's joint positions and surface from a color image. The evaluation compares predictions with reference hand geometry.

Released
September 2019
Kind
Spatial reasoning and counting
Level
School
Scored
Checked automatically
Answers
Prediction
Tasks
All public
Measures
Capability
In and out
Images → 3D
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
32
Entries on its own leaderboard, 0 models we track

ExtPose tops the leaderboard with 4.90

Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. 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 ExtPose 4.90
  2. 2 HandOS 5.00
  3. 3 ScaleHP 5.00
  4. 4 GeoHand 5.10
  5. 5 MaskHand 5.50
  6. 6 WiLoR 5.50
  7. 7 EHPE (HRNet) 5.60
  8. 8 EHPE (ResNet50) 5.70
  9. 9 Hamba 5.70
  10. 10 MobRecon 5.70
  11. 11 HHMR 5.80
  12. 12 MeshGraphormer 5.90
  13. 13 HaMeR 6.00
  14. 14 PointHMR 6.10
  15. 15 AMVUR 6.20
  16. 16 Deformer 6.20
  17. 17 SAR 6.30
  18. 18 FastMETRO 6.50
  19. 19 METRO 6.50
  20. 20 FastViT-MA36 6.60
  21. 21 PeCLR 6.60
  22. 22 I2UV-HandNet 6.70
  23. 23 CMR 6.90
  24. 24 HIU-DMTL 7.10
  25. 25 I2L-MeshNet 7.40
  26. 26 Hand4Whole 7.70
  27. 27 Pose2Mesh 7.70
  28. 28 YoutubeHand 8.40
  29. 29 MANO CNN 11.0
  30. 30 PIXIE hand expert 12.0
  31. 31 ExPose (hand sub-network h) 12.2
  32. 32 DetNet 15.7

Source: Official FreiHAND leaderboard

#EntryScoreSettingDateLinkSource
1 ExtPose 4.90 opencodepapers-b7572d.gitlab.io Official leaderboard
2 HandOS 5.00 opencodepapers-b7572d.gitlab.io Official leaderboard
3 ScaleHP 5.00 opencodepapers-b7572d.gitlab.io Official leaderboard
4 GeoHand 5.10 opencodepapers-b7572d.gitlab.io Official leaderboard
5 MaskHand 5.50 opencodepapers-b7572d.gitlab.io Official leaderboard
6 WiLoR 5.50 opencodepapers-b7572d.gitlab.io Official leaderboard
7 EHPE (HRNet) 5.60 opencodepapers-b7572d.gitlab.io Official leaderboard
8 EHPE (ResNet50) 5.70 opencodepapers-b7572d.gitlab.io Official leaderboard
9 Hamba 5.70 opencodepapers-b7572d.gitlab.io Official leaderboard
10 MobRecon 5.70 opencodepapers-b7572d.gitlab.io Official leaderboard
11 HHMR 5.80 opencodepapers-b7572d.gitlab.io Official leaderboard
12 MeshGraphormer 5.90 opencodepapers-b7572d.gitlab.io Official leaderboard
13 HaMeR 6.00 opencodepapers-b7572d.gitlab.io Official leaderboard
14 PointHMR 6.10 opencodepapers-b7572d.gitlab.io Official leaderboard
15 AMVUR 6.20 opencodepapers-b7572d.gitlab.io Official leaderboard
16 Deformer 6.20 opencodepapers-b7572d.gitlab.io Official leaderboard
17 SAR 6.30 Official leaderboard
18 FastMETRO 6.50 opencodepapers-b7572d.gitlab.io Official leaderboard
19 METRO 6.50 opencodepapers-b7572d.gitlab.io Official leaderboard
20 FastViT-MA36 6.60 opencodepapers-b7572d.gitlab.io Official leaderboard
21 PeCLR 6.60 opencodepapers-b7572d.gitlab.io Official leaderboard
22 I2UV-HandNet 6.70 opencodepapers-b7572d.gitlab.io Official leaderboard
23 CMR 6.90 opencodepapers-b7572d.gitlab.io Official leaderboard
24 HIU-DMTL 7.10 opencodepapers-b7572d.gitlab.io Official leaderboard
25 I2L-MeshNet 7.40 opencodepapers-b7572d.gitlab.io Official leaderboard
26 Hand4Whole 7.70 opencodepapers-b7572d.gitlab.io Official leaderboard
27 Pose2Mesh 7.70 opencodepapers-b7572d.gitlab.io Official leaderboard
28 YoutubeHand 8.40 opencodepapers-b7572d.gitlab.io Official leaderboard
29 MANO CNN 11.0 opencodepapers-b7572d.gitlab.io Official leaderboard
30 PIXIE hand expert 12.0 opencodepapers-b7572d.gitlab.io Official leaderboard
31 ExPose (hand sub-network h) 12.2 opencodepapers-b7572d.gitlab.io Official leaderboard
32 DetNet 15.7 opencodepapers-b7572d.gitlab.io 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 Vision

All 636 in Vision
  • MMMU In use 11.5K college-level multimodal questions across 30 subjects. Multiple-choice and open accuracy. Qwen 3.6 Plus 86%
  • MMMU-Pro In use Harder MMMU: college-level multimodal questions with more options and vision-only inputs to remove text shortcuts. GPT-5.6 Sol 84.6%
  • CharXiv In use Chart understanding on real arXiv figures. Reasoning questions require synthesizing information from complex charts. Claude Fable 5 93.5%
  • MathVista In use Math reasoning in visual contexts (charts, geometry, puzzles) from 31 datasets. Answer accuracy. Seed2.1 Pro 90.7%
  • MATH-Vision (MathVision) In use 3,040 competition math problems with visual context. Answer accuracy. Qwen 3.8 Max 97.7%
  • MMStar In use Models answer 1,500 human-curated image questions that cannot be solved without the image, across 6 core capabilities. Scored by accuracy. Qwen 3.8 Max 85.9%

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 official leaderboard (32 entries).
  • Pass Models scored 32 entries on its leaderboard, none a model we track.
  • Note Newest score No scores yet.
  • Note Frontier models No models scored yet.
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on opencodepapers-b7572d.gitlab.io: 32 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