AI Paper Atlas

Research map / Video / Action recognition

Online Temporal Action Localization with Memory-Augmented Transformer

Youngkil Song, Dongkeun Kim, Minsu Cho, Suha Kwak

ECCV 2024 · Video › Action recognition · approach: Weakly supervised temporal action localization · 6 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 two separate decoders for start and end prediction instead of one (ablation)MATR · THUMOS14average mAP: 42.7 -> 49.5Improved
use MATR flag-token memory instead of MAT or E2E-LOAD memory modules (ablation)MATR · THUMOS14average mAP: 46.9 / 47.9 -> 49.5Improved
add memory queue (ablation)MATR · THUMOS14 / MUSESaverage mAPImproved
reduce segment size from 64 to 8 (ablation)OAT-OSN vs MATR · THUMOS14average mAP: OAT-OSN 44.6 -> 25.8; MATR drop only 9.1%pHurt
add region classification for end prediction (ablation)MATR end head · THUMOS14average mAPHurt
increase instance queries N from 10 to 30 (ablation)MATR · THUMOS14average mAP: 49.5 -> 46.0Hurt
add anticipation region (Ta from 0 to 16) for predicting not-yet-ended actions (ablation)MATR · THUMOS14average mAP: 40.2 -> 49.5Improved
NMS threshold far from 0.3 (0.1 or 0.7) (ablation)MATR · THUMOS14average mAP: 49.5 -> 46.2 at 0.7Hurt
reduce segment length to 16 for faster inference (ablation)MATR · THUMOS14average mAP and inference time: 49.5 -> 46.3 mAP; 167.1ms -> 53.8msMixed
larger memory queue size (ablation)MATR · THUMOS14 / MUSESaverage mAPMixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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