AI Paper Atlas

Research map / Video / Video segmentation and tracking

EvSign: Sign Language Recognition and Translation with Streaming Events

Pengyu Zhang, Hao Yin, Zeren Wang, Wenyue Chen, Sheng Ming Li, Dong Wang, Huchuan Lu, Xu Jia

ECCV 2024 · Video › Video segmentation and tracking · approach: Event-based video reconstruction · 14 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 event streams instead of RGB frames for existing SLR methods on real EvSign dataVAC, CorrNet · EvSign testWER (%): CorrNet 32.04 -> 29.95; VAC 30.43 -> 30.71Improved
use simulated events instead of RGB framesVAC, TLP, SEN, CorrNet · PHOENIX14T testWER (%): CorrNet 20.50 -> 24.55Hurt
add sparse backbone (SConv) to event baseline (ablation)B(EV) · EvSign testWER (%): 30.08 -> 30.01; FLOPS 238.88G -> 1.22GNo clear effect
add Local Token Fusion (ablation)B(EV)+SConv · EvSign testWER (%): 30.01 -> 29.48Improved
add Gloss-Aware Temporal Aggregation replacing BiLSTM (ablation)B(EV)+SConv · EvSign testWER (%): 30.01 -> 28.89Improved
add token aggregation before temporal modeling (ablation)Ours S2G · EvSign testWER (%): 40.34 -> 28.69Improved
average pooling instead of max pooling for token aggregation (ablation)Ours S2G · EvSign testWER (%): 33.24 -> 31.56Improved
LTF instead of 1D-CNN local aggregation (ablation)Ours S2G · EvSign testWER (%): 28.89 -> 28.69Improved
add gloss-aware mask attention (rho and delta) to global self-attention (ablation)Ours S2G · EvSign testWER (%): 31.14 -> 28.69Improved
soft mask instead of binarized hard mask in GAMA (ablation)Ours S2G · EvSign testWER (%): 29.31 -> 28.69Improved
use event input instead of RGB for SLT methodsSLT, VAC+TH, CorrNet+TH · EvSign testROUGE-L: SLT 40.05 -> 41.54; VAC+TH 39.08 -> 39.48Mixed
proposed model versus RGB SLT on synthetic PHOENIX14TOurs S2GT · PHOENIX14T testROUGE-L: 40.26 (CorrNet+TH RGB) vs 40.21No clear effect

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