AI Paper Atlas

Research map / 3D vision / Human pose estimation

VPN++: Rethinking Video-Pose Embeddings for Understanding Activities of Daily Living

Srijan Das, Rui Dai, Di Yang, François Brémond

TPAMI 2021 · 3D vision › Human pose estimation · approach: Temporal video pose estimation · 98 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
replace MSE with supervised contrastive distillation (SCD) loss (ablation)VPN-F · Smarthome CStop-1 accuracy: 61.8 -> 67.1Improved
replace CRD with SCD loss (ablation)VPN-F · Smarthome CStop-1 accuracy: 64.7 -> 67.1Improved
replace CRD with SCD loss (ablation)VPN-F · NTU-60 CStop-1 accuracy: 90.1 -> 90.8Improved
use pose-driven attention instead of RGB Non-Local attention (ablation)I3D · Smarthome CStop-1 accuracy: 53.6 -> 65.2Improved
add RGB Non-Local attention to I3D (ablation)I3D · Smarthome CStop-1 accuracy: 53.4 -> 53.6No clear effect
use MSE on attention weights instead of SCD at feature level (ablation)VPN-A · Smarthome CStop-1 accuracy: 55.7 -> 62.1Improved
add collaborative training of teacher and student for MSE attention distillation (ablation)VPN-A · Smarthome CStop-1 accuracy: 61.1 -> 62.1Improved
add collaborative training for SCD attention-weight distillation (ablation)VPN-A · Smarthome CStop-1 accuracy: 54.2 -> 53.1Hurt
combine VPN-F and VPN-A in one multi-teacher model instead of sequentially (ablation)VPN++ · Smarthome CStop-1 accuracy: 62.3 -> 69.0Improved
use multi-teacher VPN++ instead of late fusion of VPN-F and VPN-A (ablation)VPN++ · Smarthome CStop-1 accuracy: 68.7 -> 69.0Improved
add spatial embedding L_e to the attention teacher (ablation)VPN++ · Smarthome CStop-1 accuracy: 68.9 -> 69.0 without L_eNo clear effect
distill poses into RGB (VPN++) instead of using poses at test time (VPN), poor pose quality (ablation)VPN++ · Smarthome CS (low-quality poses)top-1 accuracy: 62.1 -> 66.8Improved

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