Research map / Label-efficient and robust learning / Recognition and long-tail
Doing the Best We Can With What We Have: Multi-Label Balancing With Selective Learning for Attribute Prediction
Emily Hand, Carlos Castillo, Rama Chellappa
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Add Selective Learning with train target distribution to AttCNN (ablation) | AttCNN · CelebA test | average accuracy: 90.97 -> 91.05 | Improved |
| Use test distribution as Selective Learning target instead of train (ablation) | AttCNN · CelebA test | average accuracy: 91.05 -> 91.07 | No clear effect |
| Use balanced target in Selective Learning (ablation) | AttCNN · CelebA test | average accuracy: 90.97 -> 85.05 | Hurt |
| Add Selective Learning with train target distribution (ablation) | AttCNN · LFWA | average accuracy: 71.21 -> 71.49 | Improved |
| Use balanced target Selective Learning (ablation) | AttCNN · LFWA | average accuracy: 71.21 -> 73.03 | Improved |
| Use balanced target Selective Learning (ablation) | AttCNN · UMD-AED | average accuracy: Table 6: 66.85 -> 59.46 (Balanced); text: 4% over AttCNN, 11% over MOON | Mixed |
| Use P(a)=0.5 target in Selective Learning (ablation) | AttCNN · UMD-AED | average accuracy: 66.85 -> 71.11 | Improved |
| Use train-distribution Selective Learning (ablation) | AttCNN · UMD-AED | average accuracy: 66.85 -> 67.40 | Improved |
| Replace MOON per-sample reweighting with batch-level balancing (Selective Learning) | attribute CNN · CelebA test | average accuracy: MOON Balanced 86.33 vs AttCNN Balanced 85.05 on CelebA; 70.49 vs 73.03 on LFWA | Mixed |
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.
- MOON — MOON: A Mixed Objective Optimization Network for the Recognition of Facial Attributes (ECCV 2016) · better on 1 of 1 shared result
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
- MOON: A Mixed Objective Optimization Network for the Recognition of Facial Attributes (ECCV 2016 · similarity 0.89)
- Attributes for Improved Attributes: A Multi-Task Network Utilizing Implicit and Explicit Relationships for Facial Attribute Classification (AAAI 2017 · similarity 0.87)
- Heterogeneous Face Attribute Estimation: A Deep Multi-Task Learning Approach (TPAMI 2017 · similarity 0.85)
- Pedestrian Attribute Recognition as Label-balanced Multi-label Learning (ICML 2024 · similarity 0.84)
- Attributes Shape the Embedding Space of Face Recognition Models (ICML 2025 · similarity 0.83)
- Distributionally Generative Augmentation for Fair Facial Attribute Classification (CVPR 2024 · similarity 0.83)
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.
