Research map / Label-efficient and robust learning / Recognition and long-tail
Fine-Grained Fashion Representation Learning by Online Deep Clustering
Yang Jiao, Ning Xie, Yan Gao, Chien-chih Wang, Yi Sun
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 cluster-level prototypical triplet loss (ablation) | ASEN baseline · FashionAI (10% labeled) | MAP@all: 41.35 -> 42.34 | Improved |
| add cluster-level prototypical triplet loss (ablation) | ASEN++ baseline · FashionAI (100% labeled) | MAP@all: 64.27 -> 65.07 | Improved |
| add image augmentation (self-supervised constraint) (ablation) | MODC(ASEN) with cluster-level loss · FashionAI (10% labeled) | MAP@all: 42.34 -> 44.85 | Improved |
| add pseudo-label assignment for unlabeled data (ablation) | MODC(ASEN) with augmentation · FashionAI (10% labeled) | MAP@all: 44.85 -> 46.25 | Improved |
| add top1 class-specific space segmentation (ablation) | MODC(ASEN) full training · FashionAI (10% labeled) | MAP@all: 46.25 -> 56.64 | Improved |
| add top2 class-specific space segmentation (ablation) | MODC(ASEN) full training · FashionAI (10% labeled) | MAP@all: 46.25 -> 51.99 | Improved |
| add top1 segmentation to supervised MODC (ablation) | MODC(ASEN++) · FashionAI (100% labeled) | MAP@all: 65.07 -> 74.32 | Improved |
| use top1 instead of top2 segmentation (ablation) | MODC(ASEN++) · FashionAI (100% labeled) | MAP@all: 68.75 -> 74.32 | Improved |
| use top2 instead of top1 segmentation (ablation) | MODC(ASEN) · DARN (10% labeled) | MAP@all: 53.21 -> 53.55 | Mixed |
| apply MODC top1 on ASEN | ASEN · FashionAI (10% labeled) | MAP@all: 41.35 -> 56.64 | Improved |
| apply MODC top1 on ASEN++ | ASEN++ · DARN (100% labeled) | MAP@all: 55.78 -> 62.56 | Improved |
| apply MODC top1 on ASENv2 | ASENv2 · DARN (10% labeled) | MAP@all: 50.02 -> 57.60 | Improved |
Closest papers by meaning
- Learning Attribute and Class-Specific Representation Duet for Fine-Grained Fashion Analysis (CVPR 2023 · similarity 0.93)
- Fine-Grained Fashion Similarity Learning by Attribute-Specific Embedding Network (AAAI 2020 · similarity 0.91)
- Fashionpedia: Ontology, Segmentation, and an Attribute Localization Dataset (ECCV 2020 · similarity 0.88)
- Learning Attribute-Driven Disentangled Representations for Interactive Fashion Retrieval (ICCV 2021 · similarity 0.87)
- Easy Regional Contrastive Learning of Expressive Fashion Representations (NeurIPS 2024 · similarity 0.86)
- FashionERN: Enhance-and-Refine Network for Composed Fashion Image Retrieval (AAAI 2024 · similarity 0.85)
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.
