Research map / Image generation / Editing and style transfer
Face Image Retrieval With Attribute Manipulation
Alireza Zaeemzadeh, Shabnam Ghadar, Baldo Faieta, Zhe Lin, Nazanin Rahnavard, Mubarak Shah, Ratheesh Kalarot
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 L1 sparsity (lambda 1e-3) to orthogonal attribute directions (ablation) | Ours · CelebA sparsity ablation (top-5, 1000 queries) | nDCG: 0.826 -> 0.847 | Improved |
| raise sparsity lambda from 1e-3 to 5e-3 (ablation) | Ours · CelebA sparsity ablation (top-5, 1000 queries) | nDCG: 0.847 -> 0.858 | Improved |
| raise sparsity lambda from 5e-3 to 1e-2 (ablation) | Ours · CelebA sparsity ablation (top-5, 1000 queries) | nDCG: 0.858 -> 0.849 | Hurt |
| add sparsity (lambda 5e-3) on identity similarity (ablation) | Ours · CelebA sparsity ablation (top-5, 1000 queries) | identity similarity: 0.863 -> 0.864 | No clear effect |
| enforce orthogonality only (lambda=0) vs InterFaceGAN (ablation) | InterFaceGAN · CelebA sparsity ablation (top-5, 1000 queries) | identity similarity: 0.859 -> 0.863 | Improved |
| learn sparse orthogonal directions vs InterFaceGAN, no preference | InterFaceGAN · CelebA attribute-guided retrieval (1000 queries, single attribute) top-5 | nDCG: 0.568 -> 0.595 | Improved |
| learn sparse orthogonal directions vs InterFaceGAN, identity constrained | InterFaceGAN · CelebA attribute-guided retrieval (1000 queries, single attribute) top-5 | nDCG: 0.822 -> 0.858 | Improved |
| learn sparse orthogonal directions vs InterFaceGAN, best nDCG | InterFaceGAN · CelebA attribute-guided retrieval (1000 queries, single attribute) top-5 | nDCG: 0.905 -> 0.923 | Improved |
| assign per-query attribute preference (Best nDCG) instead of no preference (ablation) | Ours · CelebA attribute-guided retrieval (1000 queries, single attribute) top-5 | nDCG: 0.595 -> 0.923 | Improved |
| increase preference for the target attribute (ablation) | Ours · CelebA attribute-guided retrieval (1000 queries, single attribute) top-5 | identity similarity: 0.84 -> 0.82 | Hurt |
| GAN latent method with no preference vs compositional learning TIRG++ | TIRG++ · CelebA attribute-guided retrieval (1000 queries, single attribute) top-5 | nDCG: 0.822 -> 0.595 | Hurt |
Closest papers by meaning
- Retrieve in Style: Unsupervised Facial Feature Transfer and Retrieval (ICCV 2021 · similarity 0.88)
- Robust RGB-D Face Recognition Using Attribute-Aware Loss (TPAMI 2019 · similarity 0.85)
- Heterogeneous Face Attribute Estimation: A Deep Multi-Task Learning Approach (TPAMI 2017 · similarity 0.85)
- Attributes Shape the Embedding Space of Face Recognition Models (ICML 2025 · similarity 0.84)
- Learning Attribute-Driven Disentangled Representations for Interactive Fashion Retrieval (ICCV 2021 · similarity 0.84)
- Attribute-Preserving Face Dataset Anonymization via Latent Code Optimization (CVPR 2023 · 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.
