AI Paper Atlas

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

GridFace: Face Rectification via Learning Local Homography Transformations

Erjin Zhou, Zhimin Cao, Jian Sun

ECCV 2018 · Label-efficient and robust learning › Recognition and long-tail · approach: Fair and quality-aware face recognition · 37 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
add 8x8 local homography rectification (Grid-8) to the recognition baselinebaseline · SNFace test setverification accuracy at FAR=1e-4: 63.41 -> 72.71Improved
increase homography cells from global (Grid-1) to 4x4 (Grid-4) to 8x8 (Grid-8) (ablation)GridFace · SNFace test setverification accuracy at FAR=1e-4: 66.15 -> 68.70 -> 71.09 -> 72.71Improved
add face prior regularization to Grid-8 (ablation)Grid-8 · SNFace test setverification accuracy at FAR=1e-4: 69.05 -> 72.71Improved
add face prior regularization to Grid-8 (ablation)Grid-8 · SNFace test setverification accuracy at FAR=1e-3: 85.44 -> 87.81Improved
learn rectification and recognition jointly instead of sequentially (ablation)GridFace sequential · SNFace test setverification accuracy at FAR=1e-4: 41.6 -> 72.71Improved
train recognition on 3D-frontalized faces (baseline-3D) instead of original facesbaseline · SNFace test setverification accuracy at FAR=1e-4: 63.41 -> 58.20Hurt
train recognition on 3D-frontalized faces (baseline-3D) instead of original facesbaseline · SNFace test setverification accuracy at FAR=1e-2: 92.94 -> 94.02Improved
replace 3D frontalization (baseline-3D) with Grid-8baseline-3D · SNFace test setverification accuracy at FAR=1e-4: 58.20 -> 72.71Improved
add Grid-8 rectificationBaseline (SNFace training) · LFWopen-set DIR at FAR 1%: 80.3 -> 94.1Improved
add Grid-8 rectificationBaseline (SNFace training) · LFWmean accuracy: 99.15 -> 99.70Improved
add Grid-8 rectificationBaseline (SNFace training) · YTFmean accuracy: 94.0 -> 95.6Improved
add Grid-8 rectificationBaseline (SNFace training) · Multi-PIE yaw 90identification rate: 65.5 -> 75.4Improved

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