Research map / Label-efficient and robust learning / Recognition and long-tail
Fashion Outfit Complementary Item Retrieval
Yen-Liang Lin, Son Tran, Larry S. Davis
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 CSA-Net with outfit ranking loss instead of Type-aware | Fashion outfit compatibility and FITB · Polyvore Outfits-D | FITB Accuracy: 55.65 -> 59.26 | Improved |
| use CSA-Net with outfit ranking loss instead of Type-aware | Fashion outfit compatibility and FITB · Polyvore Outfits-D | Compat. AUC: 0.84 -> 0.87 | Improved |
| use CSA-Net with outfit ranking loss instead of SCE-Net average | Fashion outfit compatibility and FITB · Polyvore Outfits | FITB Accuracy: 59.07 -> 63.73 | Improved |
| use CSA-Net with outfit ranking loss instead of SCE-Net average | Fashion outfit compatibility and FITB · Polyvore Outfits | Compat. AUC: 0.88 -> 0.91 | Improved |
| use CSA-Net with outfit ranking loss instead of SCE-Net average | Complementary item retrieval · Polyvore Outfits-D | Recall@Top 10: 4.41% -> 5.93% | Improved |
| use CSA-Net with outfit ranking loss instead of SCE-Net average | Complementary item retrieval · Polyvore Outfits | Recall@Top 10: 5.10% -> 8.27% | Improved |
| add outfit-level ranking loss instead of triplet loss (ablation) | CSA-Net · Polyvore Outfits-D / Polyvore Outfits | FITB Accuracy: 56.17 / 60.91 -> 59.26 / 63.73 | Improved |
| add outfit-level ranking loss instead of triplet loss (ablation) | CSA-Net · Polyvore Outfits-D / Polyvore Outfits | Compat. AUC: 0.85 / 0.90 -> 0.87 / 0.91 | Improved |
| use min instead of average for negative aggregation (ablation) | CSA-Net outfit ranking loss · Polyvore Outfits-D / Polyvore Outfits | FITB Accuracy: 56.19 / 60 -> 59.26 / 63.73 | Improved |
| use min instead of average for negative aggregation (ablation) | CSA-Net outfit ranking loss · Polyvore Outfits-D / Polyvore Outfits | Compat. AUC: 0.84 / 0.89 -> 0.87 / 0.91 | Improved |
| flip the order of input categories during training (ablation) | CSA-Net outfit ranking loss · Polyvore Outfit | model performance | No clear effect |
Closest papers by meaning
- Personalized Outfit Recommendation With Learnable Anchors (CVPR 2021 · similarity 0.88)
- Learning Type-Aware Embeddings for Fashion Compatibility (ECCV 2018 · similarity 0.87)
- TransNFCM: Translation-Based Neural Fashion Compatibility Modeling (AAAI 2019 · similarity 0.87)
- Learning Attribute-Driven Disentangled Representations for Interactive Fashion Retrieval (ICCV 2021 · similarity 0.86)
- Deciphering Compatibility Relationships with Textual Descriptions via Extraction and Explanation (AAAI 2024 · similarity 0.84)
- FashionERN: Enhance-and-Refine Network for Composed Fashion Image Retrieval (AAAI 2024 · 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.
