AI Paper Atlas

Research map / Video / Action recognition

RESOUND: Towards Action Recognition without Representation Bias

Yingwei Li, Yi Li, Nuno Vasconcelos

ECCV 2018 · Video › Action recognition · approach: Disentangled action recognition · 358 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
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · KTH (K=3 subset)object bias (lower is better): 0.99 ± 0.09 -> 0.39Improved
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · Hollywood2 (K=3 subset)object bias (lower is better): 0.86 ± 0.07 -> 0.44Improved
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · HMDB51 (K=3 subset)object bias (lower is better): 0.82 ± 0.54 -> 0.00Improved
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · UCF101 (K=3 subset)object bias (lower is better): 1.08 ± 0.02 -> 0.55Improved
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · ActivityNet (K=3 subset)object bias (lower is better): 0.89 ± 0.10 -> 0.41Improved
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · Kinetics (K=3 subset)object bias (lower is better): 1.00 ± 0.11 -> 0.41Improved
choose 3 classes by explicit RESOUND optimization instead of random sampling3-class dataset sampling · Charades (K=3 subset)object bias (lower is better): 0.62 ± 0.20 -> 0.00Improved
increase temporal extent of C3D from 8 to 64 frames (ablation)C3D · Diving48accuracy (%): 11.51 -> 27.60Improved
add optical flow stream to TSN RGBTSN · Diving48accuracy (%): 16.77 -> 20.28Improved
use Diving48 instead of Kinetics as datasetResNet50 static features · representation bias (object)B(D, phi_object): 4.51 -> 1.48Improved
increase dataset size (Kinetics vs smaller datasets)action recognition datasets · object biasB(D, phi_object): 3.16 -> 4.51Hurt

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