AI Paper Atlas

Research map / Video / Video-language understanding

FingerCap: Fine-grained Finger-level Hand Motion Captioning

Xin Shen, Rui Zhu, Lei Shen, Zhuojie Wu, Xinyu Wang, Kaihao Zhang, Tianqing Zhu, Shuchen Wu, Chenxi Miao, Weikang Li, Deguo Xia, Jizhou Huang and 1 more

ECCV 2026 · Video › Video-language understanding · approach: Egocentric video understanding · 3 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add FiGOP pose stream and motion-aware projector over plain MM Projector SFTQwen3-VL-8B-Instruct · FingerCap-40K averageCIDEr: 107.28 -> 155.27Improved
add FiGOP over MM Projector SFT, HandJudgeQwen3-VL-8B-Instruct · FingerCap-40K averageHandJudge overall: 2.13 -> 2.74Improved
fine-tune with MM Projector instead of zero-shotQwen3-VL-8B-Instruct · FingerCap-40K averageCIDEr: 29.14 -> 107.28Improved
fine-tune with MM Projector only, HandJudge vs GeminiQwen3-VL-8B-Instruct · FingerCap-40K averageHandJudge overall: 2.13 vs 2.64 (Gemini-2.5-Pro)No clear effect
add FiGOP to Qwen2.5-VL-7BQwen2.5-VL-7B-Instruct · FingerCap-40KCIDEr: 101.45 -> 148.20Improved
add FiGOP on out-of-distribution subsetsQwen3-VL-8B-Instruct · OOD (BSL, HOI4D, MotionBench)HandJudge: 1.61 -> 2.04Improved
add FiGOP on OOD, CIDErQwen3-VL-8B-Instruct · OOD (BSL, HOI4D, MotionBench)CIDEr: 47.66 -> 52.21Improved
use 8 pose frames instead of 4 (ablation)FiGOP · FingerCap-40K averageCIDEr: 147.50 -> 155.27Improved
use 8 pose frames instead of 16 (ablation)FiGOP · FingerCap-40K averageCIDEr: 153.80 -> 155.27Improved
train on mixed Gesture+HOI instead of Gesture only, tested on HOI (ablation)FiGOP · FingerCap-40K HOICIDEr: 58.49 -> 165.31Improved
train on mixed instead of HOI only, tested on Gesture (ablation)FiGOP · FingerCap-40K GestureCIDEr: 42.10 -> 146.29Improved
train on mixed instead of Gesture only, tested on Gesture (ablation)FiGOP · FingerCap-40K GestureCIDEr: 154.50 -> 146.29Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy