AI Paper Atlas

Research map / Video / Action recognition

A Coarse-to-Fine Framework for Resource Efficient Video Recognition

Zuxuan Wu, Hengduo Li, Yingbin Zheng, Caiming Xiong, Yu–Gang Jiang, Larry Steven Davis

IJCV 2021 · Video › Action recognition · approach: Contrastive video representation learning · 44 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 synchronization of fine LSTM with coarse LSTM states (ablation)LiteEval · FCVIDmAP: 65.7% -> 80.0%Improved
use Gumbel-Softmax instead of REINFORCE for gating (ablation)LiteEval · FCVIDmAP and GFLOPs: 74.2% -> 80.0%; 245.9 -> 94.3 GFLOPsImproved
use Gumbel-Softmax instead of REINFORCE for gating (ablation)LiteEval · ActivityNetmAP and GFLOPs: 65.2% -> 72.7%; 269.3 -> 95.1 GFLOPsImproved
replace uniform sampling of 25 frames with coarse-to-fine gatingLiteEval · FCVIDGFLOPs: 195.5 -> 94.3 GFLOPs at 80.0% mAPImproved
replace uniform sampling of 25 frames with coarse-to-fine gatingLiteEval · ActivityNetmAP: 70.0% -> 72.7%; 195.5 -> 95.1 GFLOPsImproved
lower fine-feature usage target gamma from 0.05 to 0.01 (ablation)LiteEval · FCVIDmAP: 80.0% -> 78.8%; 94.3 -> 75.4 GFLOPsHurt
raise gamma from 0.05 to 0.10 (ablation)LiteEval · FCVIDmAP: 80.0% -> 80.1%; 94.3 -> 139.0 GFLOPsNo clear effect
raise gamma from 0.01 to 0.03 (ablation)LiteEval · FCVIDmAP: 78.8% -> 79.7%; 75.4 -> 82.1 GFLOPsImproved
increase coarse LSTM hidden units from 64 to 512 (ablation)LiteEval · FCVIDmAP: 76.9% -> 80.0%Improved
increase coarse LSTM hidden units from 256 to 512 (ablation)LiteEval · FCVIDmAP: 78.3% -> 80.0%Improved
compare with FrameGlimpse and FastForward agents that jump to future framesLiteEval · FCVIDmAP: 71.2% -> 80.0%; 67.6% -> 80.0%Improved

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