AI Paper Atlas

Research map / Video / Action recognition

TravelNet: Self-Supervised Physically Plausible Hand Motion Learning From Monocular Color Images

Zimeng Zhao, Xi Zhao, Yangang Wang

ICCV 2021 · Video › Action recognition · approach: Skeleton-based action recognition · 15 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 physical pose representation θ to the pose estimator (ablation)TravelNet hand pose estimator · DOAUC of PCK: .940 -> .962Improved
add physical pose representation θ to the pose estimator (ablation)TravelNet hand pose estimator · STBAUC of PCK: .890 -> .998Improved
add unified bone ratio γ extraction (ablation)TravelNet hand pose estimator · RHDAUC of PCK: .861 -> .903Improved
add self-supervised FKLayer term (ablation)TravelNet hand pose estimator · STBAUC of PCK: .956 -> .998Improved
add TravelNet over naive temporal smoothness of frame-wise estimates (ablation)TravelNet · Noisy reconstruction σ=1.0average 3D joint error (mm): 32.57 -> 24.30Improved
add mean-shift clustering Fϵ in the encoder (ablation)TravelNet · Noisy reconstruction σ=1.0average 3D joint error (mm): 28.05 -> 24.30Improved
add attention Fα in the encoder (ablation)TravelNet · Noisy reconstruction σ=1.0average 3D joint error (mm): 26.68 -> 24.30Improved
add non-local attention in each learning block (ablation)TravelNet · Noisy reconstruction σ=1.0average 3D joint error (mm): 25.36 -> 24.30Improved
add temporal smooth term LT in training (ablation)TravelNet · Noisy reconstruction σ=1.0average 3D joint error (mm): 25.42 -> 24.30Improved
add collision penalty LC in training (ablation)TravelNet · Noisy reconstruction σ=1.0average 3D joint error (mm): 24.31 -> 24.30No clear effect
add collision penalty LC in training (ablation)TravelNet · DO sequencesaverage 3D joint error (mm): 17.05 -> 17.04No clear effect
use TravelNet fine-tuned encoder-decoder instead of frame-by-frame estimationTravelNet · DO sequences (severe occlusion)average 3D joint error (mm): 22.57 -> 17.04Improved

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