Research map / Vision-language and multimodal / CLIP and zero-shot
FashionViL: Fashion-Focused Vision-and-Language Representation Learning
Xiao Han, Licheng Yu, Xiatian Zhu, Li Zhang, Yi-Zhe Song, Tao Xiang
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 multimodal pre-training (ablation) | FashionViL · FashionGen test (1K retrieval) | ITR R@1: 58.84 -> 65.54 | Improved |
| unfreeze image encoder for end-to-end training (ablation) | FashionViL (no pre-training) · FashionGen test (1K retrieval) | ITR R@1: 21.13 -> 58.84 | Improved |
| add multimodal pre-training (ablation) | FashionViL · FashionIQ (end-to-end ResNet50) | mean recall: 39.67 -> 44.12 | Improved |
| add multimodal pre-training (ablation) | FashionViL · FashionIQ (fixed ResNet152) | mean recall: 31.78 -> 34.19 | Improved |
| add multimodal pre-training (image encoder only) (ablation) | FashionViL · PolyvoreOutfits disjoint (OCIR) | mean recall: 9.90 -> 11.98 | Improved |
| add multimodal pre-training (ablation) | FashionViL · FashionGen category/subcategory recognition (SCR) | accuracy: 91.45 -> 92.23 | Improved |
| add PAC pretraining to the other five tasks (ablation) | FashionViL · Pre-training task ablation (5 downstream tasks) | Meta-sum: 277.84 -> 279.36 | Improved |
| add MVC to ITC+MLM+MPFC+ITM (ablation) | FashionViL · Pre-training task ablation (5 downstream tasks) | Meta-sum: 276.62 -> 277.84 | Improved |
| add ITM to ITC+MLM+MPFC (ablation) | FashionViL · Pre-training task ablation (5 downstream tasks) | Meta-sum: 275.94 -> 276.62 | Improved |
| add MLM and MPFC to ITC (ablation) | FashionViL · Pre-training task ablation (5 downstream tasks) | Meta-sum: 275.31 -> 275.94 | Mixed |
| use multi-angle images in MVC rather than augmented same image (ablation) | MVC · Pre-training task ablation (5 downstream tasks) | Meta-sum: 266.32 -> 269.38 | Improved |
| apply pseudo-attribute supervision on global representation (PAC) instead of masked-token MLM (ablation) | PAC · Pre-training task ablation (5 downstream tasks) | Meta-sum: 266.92 -> 269.07 | Improved |
Later papers that compared with this method
3 later accepted papers report FashionViL as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Integration of Global and Local Representations for Fine-grained Cross-modal Alignment (ECCV 2024 · better on 3 of 3 shared results)
- FAME-ViL: Multi-Tasking Vision-Language Model for Heterogeneous Fashion Tasks (CVPR 2023 · better on 2 of 2 shared results)
- FashionSAP: Symbols and Attributes Prompt for Fine-Grained Fashion Vision-Language Pre-Training (CVPR 2023 · better on 2 of 2 shared results)
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.
- KaleidoBERT — Kaleido-BERT: Vision-Language Pre-Training on Fashion Domain (CVPR 2021) · better on 5 of 5 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
- Easy Regional Contrastive Learning of Expressive Fashion Representations (NeurIPS 2024 · similarity 0.92)
- Seeing the Abstract: Translating the Abstract Language for Vision Language Models (CVPR 2025 · similarity 0.90)
- FashionERN: Enhance-and-Refine Network for Composed Fashion Image Retrieval (AAAI 2024 · similarity 0.88)
- SyncMask: Synchronized Attentional Masking for Fashion-centric Vision-Language Pretraining (CVPR 2024 · similarity 0.87)
- VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks (ICLR 2025 · similarity 0.87)
- MMRL: Multi-Modal Representation Learning for Vision-Language Models (CVPR 2025 · similarity 0.86)
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.
