AI Paper Atlas

Research map / Video / Action recognition

Learning State-Aware Visual Representations from Audible Interactions

Himangi Mittal, Pedro Morgado, Unnat Jain, Abhinav Gupta

NeurIPS 2022 · Video › Action recognition · approach: Contrastive video representation learning · 13 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
sample training clips around audio-detected moments of interaction instead of random times (ablation)RepLAI · EPIC-Kitchens-100 action recognitionverb top-1 accuracy: 28.71 -> 31.71Improved
sample training clips around audio-detected moments of interaction instead of random times (ablation)RepLAI · Ego4D state change classificationaccuracy: 62.90 -> 66.30Improved
sample training clips around audio-detected moments of interaction instead of random times (ablation)RepLAI · Ego4D long-term action anticipationverb ED@(Z=20) (lower better): 0.792 -> 0.755Improved
add the AStC audible state-change loss to AVC (ablation)RepLAI · EPIC-Kitchens-100 action recognitionverb top-1 accuracy: 29.29 -> 31.71Improved
add the AStC audible state-change loss to AVC (ablation)RepLAI · Ego4D state change classificationaccuracy: 63.60 -> 66.30Improved
add the AVC loss to AStC (ablation)RepLAI · EPIC-Kitchens-100 action recognitionverb top-1 accuracy: 29.92 -> 31.71Improved
add the AVC loss to AStC (ablation)RepLAI · Ego4D state change classificationaccuracy: 64.00 -> 66.30Improved
initialize from AVID audio-visual pretraining instead of scratch (ablation)RepLAI · EPIC-Kitchens-100 action recognitionverb top-1 accuracy: 25.75 -> 31.71Improved
initialize from AVID audio-visual pretraining instead of scratch (ablation)RepLAI · Ego4D state change classificationaccuracy: 66.20 -> 66.30No clear effect
further train AVID with RepLAI (MoI + AVC + AStC)AVID · EPIC-Kitchens-100 action recognitionverb top-1 accuracy: 26.62 -> 31.71Improved
further train AVID with RepLAIAVID · Ego4D long-term action anticipationverb ED@(Z=20) (lower better): 0.811 -> 0.755Improved
RepLAI vs AVID on tail classesRepLAI · EPIC-Kitchens-100 tail classesverb top-1 accuracy: 16.80 -> 20.05Improved

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