Research map / Image generation / Autoregressive and latent generation
Deep Factorised Inverse-Sketching
Kaiyue Pang, Da Li, Jifei Song, Yi-Zhe Song, Tao Xiang, Timothy M. Hospedales
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 attribute prediction regularisation to the embedding subnet (ablation) | style transfer model · QMUL-Shoe-V2 retrieval with synthesised sketches | acc@1: 6.46% -> 8.26% | Improved |
| add attribute prediction regularisation to the embedding subnet (ablation) | style transfer model · QMUL-Shoe-V2 retrieval with synthesised sketches | acc@10: 31.38% -> 35.14% | Improved |
| add attribute prediction regularisation (ablation) | style transfer model · Human study (0.9, 0.1) | % preferring Ours-full: 62.0% | Improved |
| replace visual consistency with cyclic embedding consistency (vs UNIT) (ablation) | style transfer model · QMUL-Shoe-V2 retrieval with synthesised sketches | acc@1: 4.50% -> 8.26% | Improved |
| replace learned encoder with fixed VGG encoder in UNIT (UNIT-vgg) (ablation) | UNIT · QMUL-Shoe-V2 retrieval with synthesised sketches | acc@1: 4.50% -> 4.95% | Improved |
| Ours-full preferred over UNIT by humans (ablation) | style transfer model · Human study (0.9, 0.1) | % preferring Ours-full: 88.0% | Improved |
| add decorrelation loss between sketch and contour features (ablation) | four-branch FG-SBIR · QMUL-Shoe-V2 FG-SBIR | acc@1: 33.93% -> 35.89% | Improved |
| add contour-detail factorisation (four-branch with decorrelation) vs Vanilla-triplet (ablation) | FG-SBIR ResNet-50 · QMUL-Shoe-V2 FG-SBIR | acc@1: 33.48% -> 35.89% | Improved |
| without decorrelation, four-branch model vs Vanilla-triplet (ablation) | four-branch FG-SBIR · QMUL-Shoe-V2 FG-SBIR | acc@1: 33.48% -> 33.93% | No clear effect |
Closest papers by meaning
- Learning to Sketch With Shortcut Cycle Consistency (CVPR 2018 · similarity 0.92)
- Generalising Fine-Grained Sketch-Based Image Retrieval (CVPR 2019 · similarity 0.91)
- Unsupervised Sketch to Photo Synthesis (ECCV 2020 · similarity 0.91)
- Sketch Me That Shoe (CVPR 2016 · similarity 0.90)
- Picture That Sketch: Photorealistic Image Generation From Abstract Sketches (CVPR 2023 · similarity 0.90)
- Sketch Down the FLOPs: Towards Efficient Networks for Human Sketch (CVPR 2025 · similarity 0.88)
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.
