AI Paper Atlas

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

The Devil of Face Recognition is in the Noise

Fei Wang, Liren Chen, Cheng Li, Shiyao Huang, Yanjie Chen, Chen Qian, Chen Change Loy

ECCV 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Fair and quality-aware face recognition · 219 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
manually clean the training set instead of using the noisy sample (ablation)A-Softmax on MS-Celeb-1M subset · MegaFace rank-1rank-1 accuracy (%): 70.56 -> 73.53Improved
manually clean the training set instead of using the noisy sample (ablation)Center Loss on MegaFace subset · MegaFace rank-1rank-1 accuracy (%): 62.55 -> 67.64Improved
manually clean the training set instead of using the noisy sample (ablation)Softmax on MegaFace subset · MegaFace rank-1rank-1 accuracy (%): 59.68 -> 62.86Improved
use cleaned subsets rather than sampled noisy subsets (average over losses and datasets) (ablation)Attention-56 face models · MegaFace rank-1rank-1 accuracy (%): 4.14%Improved
train on clean subset instead of 4x larger raw data (ablation)A-Softmax on MS-Celeb-1M · MegaFace rank-1rank-1 accuracy (%): 73.99 -> 73.53Mixed
train on IMDb-Face instead of MS-Celeb-1M(V1)A-Softmax · MegaFace rank-1rank-1 accuracy (%): 73.99 -> 84.06Improved
train on IMDb-Face instead of CASIA-WebFaceA-Softmax · MegaFace rank-1rank-1 accuracy (%): 70.89 -> 84.06Improved
use A-Softmax instead of Softmax on IMDb-FaceAttention-56 · MegaFace rank-1rank-1 accuracy (%): 74.75 -> 84.06Improved
inject label flips instead of outliers at equal noise ratio (ablation)Softmax, Center, A-Softmax on IMDb-Face · MegaFace rank-1rank-1 accuracyHurt
increase label-flip noise to high ratios (ablation)A-Softmax on IMDb-Face · MegaFace rank-1rank-1 accuracyHurt
dilute fixed clean data with equal, double or five times label-flip noise (ablation)Attention-56 on IMDb-Face subsets · MegaFace rank-1rank-1 accuracyHurt
annotate with Scheme I (draw box) rather than yes/no Scheme III (ablation)annotation scheme · user study on 30 identitiesF1 / true positive rate: 90%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