AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Evolving Losses for Unsupervised Video Representation Learning

AJ Piergiovanni, Anelia Angelova, Michael S. Ryoo

CVPR 2020 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Pretext-task visual pre-training · 141 citations (OpenAlex)

Check 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
Evolve the combination and weights of self-supervised and distillation losses using unsupervised Zipf matching rather than a randomly sampled loss (ablation)ELo · HMDB51fine-tune accuracy: 31.2 -> 67.4Improved
Add distillation of other modalities into the RGB network (ablation)ELo · HMDB51fine-tune accuracy: 53.7 -> 67.4Improved
Add distillation of other modalities into the RGB network (ablation)ELo · UCF101fine-tune accuracy: 84.2 -> 93.8Improved
Use label-guided HMDB clustering fitness instead of unsupervised Zipf-matching fitness (ablation)ELo · HMDB51fine-tune accuracy: 67.4 -> 67.8No clear effect
Use CMA-ES instead of tournament selection for loss-weight search (ablation)Loss evolution · HMDBbest loss fitness accuracy: 61.4 -> 67.4Improved
Use CMA-ES for 250 iterations rather than random search for 2000 iterations (ablation)Loss evolution · HMDBbest loss fitness accuracy: 52.4 -> 67.4Improved
Increase the amount of unlabeled video data while holding training iterations fixedELo · HMDBaccuracyImproved
Increase unlabeled data and training time by holding the number of epochs fixedELo · HMDBaccuracyImproved
Use unsupervised ELo pretraining before Kinetics fine-tuning instead of random initializationELo · Kinetics-400accuracy: 71.2 -> 74.4Improved
Use unsupervised ELo pretraining before Kinetics fine-tuning instead of Frame Shuffle pretrainingELo · Kinetics-400accuracy: 70.9 -> 74.4Improved

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