AI Paper Atlas

Research map / Video / Action recognition

Discovering Novel Actions from Open World Egocentric Videos with Object-Grounded Visual Commonsense Reasoning

Sanjoy Kundu, Shubham Trehan, Sathyanarayanan N Aakur

ECCV 2024 · Video › Action recognition · approach: Disentangled action recognition · 5 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 evidence-based CLIP object grounding via ConceptNet instead of KGL prior-only groundingKGL · GTEA GazeObject accuracy: 5.12 -> 13.07Improved
naively add CLIP grounding to KGLKGL · GTEA Gaze / GazePlusAction accuracy: 8.04 -> 8.15 on Gaze; average 2% gainNo clear effect
add posterior-based action refinement with visual-semantic groundingALGO · GTEA GazeAction accuracy: KGL 8.04 -> ALGO 17.05Improved
add LaViLa action priors to ALGOALGO · GTEA GazeActivity accuracy: 15.05 -> 22.05Improved
add LaViLa action priors to ALGO on EPIC-KitchensALGO · EPIC-Kitchens-100Object accuracy: verb improved to 12.54; object 22.84 vs LaViLa 23.25Mixed
run posterior refinement for more than two iterations (ablation)ALGO · GTEA GazeActivity accuracyNo clear effect
use ConceptNet Numberbatch instead of GloVe embeddings (ablation)ALGO action grounding · GTEA Gazeactivity accuracyImproved
use S3D-G instead of S3D visual features (ablation)ALGO action grounding · GTEA Gazeactivity accuracyNo clear effect
use CLIP + evidence grounding instead of CLIP only (ablation)ALGO object grounding · GTEA GazeObject accuracyImproved
replace dataset-derived search space with an unknown GPT-4 generated spaceALGO · GTEA GazeVerb accuracy: 14.11 -> 9.87Hurt

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