AI Paper Atlas

Research map / Video / Action recognition

Overcoming the Domain Gap in Neural Action Representations

Semih Günel, Florian Aymanns, Sina Honari, Pavan P Ramdya, Pascal Fua

IJCV 2022 · Video › Action recognition · approach: Disentangled action recognition · 2 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 swapping augmentation (ablation)contrastive neural-behavior model · MC2Pacross-subject accuracy: + 7.9Improved
add swapping augmentation (ablation)contrastive neural-behavior model · MC2Pidentity recognition accuracy: -63.0Improved
add swapping augmentation (ablation)contrastive neural-behavior model · MC2Psingle-subject accuracy: + 2.7Improved
add calcium augmentation on top of swapping (ablation)Ours · MC2Pacross-subject accuracy: + 2.7Improved
add calcium augmentation on top of swapping (ablation)Ours · MC2Pidentity recognition accuracy: +1.2No clear effect
add mix augmentation on top of swapping and calcium (ablation)Ours · MC2Pacross-subject accuracy: + 1.2Improved
add mix augmentation on top of swapping and calcium (ablation)Ours · MC2Pidentity recognition accuracy: -0.8No clear effect
add MMD or GRL domain adaptation to SimCLRSimCLR · MC2Pacross-subject accuracy (100%): 50.6 -> 53.1 (MMD), 50.6 -> 52.3 (GRL)No clear effect
add swapping augmentation to SimCLR on pose dataSimCLR · H3.6M Walking, Sitting, Posingacross-subject accuracy (100%): 55.7 -> 73.6Improved
add swapping augmentation to SimCLR on pose dataSimCLR · ECoG Moving vs Restacross-subject accuracy (100%): 72.1 -> 81.9Improved
use neural-only contrastive pretraining instead of multimodal SimCLROurs (Neural Only) · MC2Pacross-subject accuracy (100%): 50.6 -> 30.6Hurt
add domain adaptation (MMD or GRL) to SimCLRSimCLR · MC2Pidentity recognition accuracy (100%): 80.3 -> 21.2 (MMD), 80.3 -> 19.1 (GRL)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