AI Paper Atlas

Research map / Video / Action recognition

On the Utility of 3D Hand Poses for Action Recognition

Md Salman Shamil, Dibyadip Chatterjee, Fadime Sener, Shugao Ma, Angela Yao

ECCV 2024 · Video › Action recognition · approach: Skeleton-based action recognition · 19 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 multimodal tokenizer, anticipation loss and verb/object losses (ablation)HandFormer · H2Oaction accuracy: 85.95 -> 93.39Improved
add multimodal tokenizer, anticipation loss and verb/object losses (ablation)HandFormer · Assembly101action accuracy: 38.98 -> 41.06Improved
add multimodal tokenizer (ablation)HandFormer · Assembly101action accuracy: 38.98 -> 40.19Improved
add feature anticipation loss on top of multimodal tokenizer (ablation)HandFormer · Assembly101action accuracy: 40.19 -> 40.24No clear effect
add verb and object supervision losses (ablation)HandFormer · H2Oaction accuracy: 85.95 -> 90.50Improved
use 3D instead of 2D hand poses (ablation)HandFormer-B/6 · Assembly101 verb recognitionverb accuracy: 58.92 -> 63.70Improved
add a single RGB frame to hand poses (ablation)HandFormer · Assembly101action accuracy: 35.46 vs TSM 35.27Improved
increase the number of RGB framesHandFormer · Assembly101object accuracyMixed
replace ViT-g/14 with TSM-pretrained ResNet50 frame encoder (RGB-only) (ablation)HandFormer · Assembly101action accuracy: 32.07 -> 35.09Improved
replace ViT-g/14 with TSM-pretrained ResNet50 frame encoder (with pose) (ablation)HandFormer · Assembly101action accuracy: 41.06 -> 41.99No clear effect
combine 3D hand pose with a single RGB view instead of multiple viewsHandFormer · Assembly101action accuracy: 35.27 -> 41.06Improved
use 2D heatmaps from PoseConv3D instead of trajectory-based encodinghand poses · Assembly101 verb recognitionverb accuracy: 46.71 vs 58.92 (HandFormer 2D)Hurt

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