Research map / Detection and segmentation / CNN features
"Fashionformer: A Simple, Effective and Unified Baseline for Human Fashion Segmentation and Recognition"
Shilin Xu, Xiangtai Li, Jingbo Wang, Guangliang Cheng, Yunhai Tong, Dacheng Tao
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 joint attribute learning over instance segmentation baseline (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU: 30.1 -> 33.2 | Improved |
| add residual addition (RD) (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU+F1: 27.0 -> 29.2 | Improved |
| add dynamic convolution (DC) (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU+F1: 27.5 -> 29.2 | Improved |
| add multi-layer rendering (MLR) (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU+F1: 26.5 -> 29.2 | Improved |
| use I=3 setting instead of N=1 variant (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU: 29.1 -> 33.2 | Improved |
| increase MLR layers from 1 to 4 (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU+F1: 27.5 -> 29.2 | Improved |
| increase MLR layers from 2 to 4 (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU: 33.1 -> 33.2 | No clear effect |
| use individual query instead of shared query for attributes (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU+F1: 27.6 -> 29.1 | Improved |
| use individual query instead of shared query for attributes (ablation) | Fashionformer · Fashionpedia val (R50) | AP mask IoU: 33.4 -> 33.3 | No clear effect |
| Fashionformer instead of Attribute-Mask R-CNN (R50, 1x) | Attribute-Mask R-CNN · Fashionpedia val | AP mask IoU+F1: 25.5 -> 36.6 | Improved |
| Fashionformer instead of Attribute-Mask R-CNN (Swin-b, 3x) | Attribute-Mask R-CNN · Fashionpedia val | AP mask IoU+F1: 40.6 -> 46.5 | Improved |
| Fashionformer instead of K-Net with Swin-b | K-Net · Fashionpedia val (3x) | AP mask IoU: 46.4 -> 49.5 | Improved |
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.
- K-Net — K-Net: Towards Unified Image Segmentation (NeurIPS 2021) · better on 3 of 3 shared results
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
- Fashionpedia: Ontology, Segmentation, and an Attribute Localization Dataset (ECCV 2020 · similarity 0.89)
- FashionVLP: Vision Language Transformer for Fashion Retrieval With Feedback (CVPR 2022 · similarity 0.88)
- Easy Regional Contrastive Learning of Expressive Fashion Representations (NeurIPS 2024 · similarity 0.86)
- SegFace: Face Segmentation of Long-Tail Classes (AAAI 2025 · similarity 0.85)
- ChatGarment: Garment Estimation, Generation and Editing via Large Language Models (CVPR 2025 · similarity 0.85)
- 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.
