AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Pre-training Strategies and Datasets for Facial Representation Learning

Adrian Bulat, Shiyang Cheng, Jing Yang, Andrew Garbett, Enrique Sanchez, Georgios Tzimiropoulos

ECCV 2022 · Label-efficient and robust learning › Recognition and long-tail · approach: Margin-based face recognition losses · 48 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
use SwAV instead of MoCov2 for unsupervised pre-training (ablation)face ResNet-50 · 300W (2% data)NME (%): 11.9 -> 4.97Improved
use SwAV instead of MoCov2 for unsupervised pre-training (ablation)face ResNet-50 · DISFA (2% data)ICC: .280 -> .560Improved
use SwAV batch size 256 instead of 4096 (ablation)SwAV pre-training · 300W (2% data)NME (%): 4.97 vs 5.00No clear effect
pre-train on full VGG-Face instead of VGG-Face-small (ablation)unsupervised SwAV · 300W (2% data)NME: 5.46 -> 4.97Improved
pre-train on Large-Scale-Face (5M) instead of VGG-Face (ablation)unsupervised SwAV · 300W (2% data)NME: 4.97 vs 4.98No clear effect
pre-train on uncurated Flickr-Face instead of curated VGG-Face (ablation)unsupervised SwAV · 300W (100% data)NME: 3.85 vs 3.86No clear effect
unsupervised SwAV on faces instead of on ImageNet (ablation)SwAV pre-training · 300W (2% data)NME: 5.1 -> 4.97Improved
unsupervised face pre-training instead of training from scratchResNet-50 · 300W (2% data)NME: 13.52 -> 4.97Improved
unsupervised face pre-training instead of supervised ImageNet pre-trainingResNet-50 · AFLW2000-3D (2% data)NME: 6.15 -> 4.31Improved
unsupervised face pre-training instead of supervised VGG-F pre-trainingResNet-50 · IJB-B TAR@1e-4 (2% data)TAR: 0.694 -> 0.744Improved
fine-tune the whole network instead of linear adaptation (ablation)Ours (VGG-F) · DISFA (100% data)ICC: .342 -> .598Improved
unsupervised pre-trained student for self-distillation instead of supervised/scratch (ablation)self-distillation · 300W semi-supervisedNMEImproved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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