AI Paper Atlas

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

Dynamic Feature Learning for Partial Face Recognition

Lingxiao He, Haiqing Li, Qi Zhang, Zhenan Sun

CVPR 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Margin-based face recognition losses · 69 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 DFM instead of MR-CNN for partial-face matchingPartial face recognition · Partial-LFW, 1,000 classesRank-1 accuracy: 24.7% -> 27.3%Improved
Use DFM instead of the strongest listed baseline, MR-CNNPartial face recognition · CASIA-NIR-DistanceRank-1 accuracy: 85.97 -> 94.96Improved
Use DFM instead of the strongest listed baseline, MR-CNNPartial face recognition · Partial-YTFRank-1 accuracy: 57.00 -> 61.00Improved
Use DFM with spatial feature maps rather than resize partial images to fixed-size inputsPartial face recognition · Partial-LFW, 1,000 classesRank-1 accuracy: 20.9% -> 27.3%Improved
Use DFM instead of VGGFace on arbitrary partial-face probesPartial face verification · LFW verification protocolTrue positive rateImproved
Use holistic probes rather than a mixture of partial and holistic probesDFM · Partial-LFW, 1,000 classesRank-1 accuracy: 41.9% -> 56.8%Improved
Use a mixture of partial and holistic probes rather than only partial probesDFM · Partial-LFW, 1,000 classesRank-1 accuracy: 27.3% -> 41.9%Improved
Present the upper face area rather than the lower face areaDFM · Partial-LFWRank-1 accuracy: 7.8% -> 39.2%Improved
Increase the cropped patch sizeDFM · Partial-LFWRank-1 accuracyImproved
Set similarity-guided constraint weight α to 2.1 and sparsity weight β to 0.1 (ablation)DFM · Partial-LFW, 1,000 classesRank-1 accuracy: 27.3%Improved
Use larger α after its low-to-high sweep (ablation)DFM · Partial-LFWRank-1 accuracyMixed
Set sparsity weight β to 0.1 rather than the other tested values (ablation)DFM · Partial-LFWRank-1 accuracyImproved

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