Research map / Label-efficient and robust learning / Recognition and long-tail
Personalized Outfit Recommendation With Learnable Anchors
Zhi Lu, Yang Hu, Yan Chen, Bing Zeng
Check 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 |
|---|---|---|---|
| Use learnable user anchors and anchor-based preference scores | LPAE-u · Polyvore-630 | AUC: 0.8872 -> 0.9006 | Improved |
| Use learnable user anchors and anchor-based preference scores | LPAE-u · IQON-550 | NDCG: 0.8496 -> 0.8727 | Improved |
| Add shared general anchors (ablation) | LPAE-u · Polyvore-53, 5 outfits | AUC: 0.8144 -> 0.8299 | Improved |
| Add shared general anchors (ablation) | LPAE-u · Polyvore-519 | AUC: 0.9119 -> 0.9119 | No clear effect |
| Use anchor-search rather than direct optimization for the new-user profile (ablation) | LPAE-u · Polyvore-53, 1 outfit | AUC: 0.6886 ± .0039 -> .7894 ± .0011 | Improved |
| Use anchor-search rather than user-search for the new-user profile (ablation) | LPAE-u · Polyvore-53, 1 outfit | AUC: .7467 ± .0024 -> .7894 ± .0011 | Improved |
| Use anchor-search rather than direct optimization for the new-user profile (ablation) | LPAE-g · Polyvore-53, 1 outfit | AUC: .7840 ± .0023 -> .7938 ± .0009 | Improved |
| Use user-search rather than anchor-search for the new-user profile (ablation) | LPAE-u · Polyvore-53, 100 outfits | AUC: .8623 ± .0004 -> .8252 ± .0011 | Hurt |
| Represent each user with multiple anchors instead of one (ablation) | LPAE-u · Recommendation tasks | AUC | Improved |
| Increase the anchor count beyond the useful range (ablation) | LPAE-u · Recommendation tasks | AUC | Hurt |
| Use self-attention item aggregation for outfit representations | LPAE · Polyvore-519, Polyvore-630, and IQON-550 | AUC | Improved |
Closest papers by meaning
- Learning Binary Code for Personalized Fashion Recommendation (CVPR 2019 · similarity 0.89)
- Personalized Fashion Design (ICCV 2019 · similarity 0.89)
- Fashion Outfit Complementary Item Retrieval (CVPR 2020 · similarity 0.88)
- Towards Unified Multi-Modal Personalization: Large Vision-Language Models for Generative Recommendation and Beyond (ICLR 2024 · similarity 0.86)
- StyleTailor: Towards Personalized Fashion Styling via Hierarchical Negative Feedback (AAAI 2026 · similarity 0.86)
- A Geometric Approach to Personalized Recommendation with Set-Theoretic Constraints Using Box Embeddings (ICML 2025 · similarity 0.84)
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.
