AI Paper Atlas

Research map / Video / Action recognition

SAFER-Activities: A Dataset for Smart Assessment of Fall Events and Routine Activities

Diwas Lamsal, Pramod Wickramatilake, Jednipat Moonrinta, Mongkol Ekpanyapong, Matthew Dailey

ECCV 2026 · Video › Action recognition · approach: Skeleton-based action recognition · 0 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
use 2D skeleton instead of frozen RGB featuresMS-G3D / VideoMAE · SAFER-Activities In-Labmacro F1: VideoMAE 80.3 vs MS-G3D 87.2Improved
evaluate frozen RGB under domain shift (In-Lab to Non-Lab)DINOv3 / CLIP / VideoMAE · SAFER-Activitiesmacro F1: DINOv3 72.1 -> 23.8; CLIP 70.6 -> 35.6; VideoMAE 80.3 -> 48.7Hurt
evaluate skeleton model under domain shift (In-Lab to Non-Lab)PoseC3D · SAFER-Activitiesmacro F1: 86.6 -> 79.3Hurt
fuse frozen RGB with CNN1D skeleton by concatenationCNN1D · SAFER-Activities Wheelchairmacro F1: 68.2 -> 78.6Improved
fuse frozen RGB with CNN1D skeleton by concatenationCNN1D · SAFER-Activities Non-Lab OODmacro F1: 73.8 -> 32.9 (DINOv3), 54.5 (VideoMAE)Hurt
fuse frozen RGB with CNN1D skeleton by concatenationCNN1D · SAFER-Activities In-Labmacro F1: 83.8 -> 85.8 (VideoMAE)Improved
replace concatenation with modality dropout (ModDrop) (ablation)fusion · SAFER-Activities Non-Lab OODmacro F1: VideoMAE 54.5 -> 67.3; DINOv3 32.9 -> 59.6Improved
replace concatenation with QMF or OGM-GE (ablation)fusion · SAFER-Activities Non-Lab OODmacro F1: DINOv3 32.9 -> 32.5 (QMF), 29.0 (OGM-GE)No clear effect
use MMCL (skeleton only at inference) (ablation)CNN1D fusion · SAFER-Activities Wheelchairmacro F1: 68.2 -> 69.7 (CLIP)Improved
increase temporal stride to 3 (144-frame context) (ablation)2D skeleton models · SAFER-Activities Wheelchairmacro F1: ST-GCN++ 71.9 -> 75.6; PoseC3D 75.8 -> 78.3Improved
increase temporal stride to 3 (ablation)2D skeleton models · SAFER-Activities In-Labmacro F1: MS-G3D 87.2 -> 87.1No clear effect
test skeleton vs frozen RGB on ImViA fall detectionPoseC3D / CLIP / DINOv3 · ImViAfall F1: CLIP 0.0 vs PoseC3D 96.8Improved

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