Research map / Label-efficient and robust learning / Recognition and long-tail
Cross-Domain Visual Matching via Generalized Similarity Measure and Feature Learning
Liang Lin, Guangrun Wang, Wangmeng Zuo, Xiangchu Feng, Lei Zhang
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 |
|---|---|---|---|
| replace handcrafted features with end-to-end deep feature learning (ablation) | generalized similarity model · CUHK03 | rank-1 accuracy (%): 31.85 -> 58.4 | Improved |
| replace handcrafted features with end-to-end deep feature learning (ablation) | generalized similarity model · CUHK01 | rank-1 accuracy (%): 39.5 -> 66.5 | Improved |
| add affine transformation to the similarity model (ablation) | deep generalized similarity · CUHK03 | rank-1 accuracy (%) | Improved |
| use generalized similarity instead of LDM on handcrafted features (ablation) | fixed handcrafted features · CUHK03 | rank-1 accuracy (%): 13.51 -> 31.85 | Improved |
| use sample-based batch instead of sample-pair-based batch (ablation) | training · CUHK03 | rank-1 accuracy (%): 46.96 -> 58.14 | Improved |
| use generalized similarity measure and CNN features instead of IDLA | person re-id · CUHK03 | rank-1 accuracy (%): 54.74 -> 58.39 | Improved |
| add more training individuals (Ours-2, 500 vs 200 individuals) | Ours · CACD retrieval | retrieval rate | Improved |
| use generalized similarity measure instead of CARC | age-invariant face · CACD-VS | verification accuracy (%): 87.6 -> 89.8 | Improved |
| use generalized similarity measure instead of CARC | age-invariant face · MORPH | recognition rate (%): 92.8 -> 94.4 | Improved |
| use generalized similarity instead of PSCL-EA | still-video face · COX V2S | recognition rate (%): 30.33 -> 28.45 | Hurt |
| use generalized similarity instead of PSCL-EA | still-video face · COX S2V | recognition rate (%): 28.39 -> 29.02 | Improved |
| use generalized similarity instead of MRF+ | sketch-photo · CUFS | recognition rate (%): 99.0 -> 100.0 | Improved |
Closest papers by meaning
- Similarity Learning With Spatial Constraints for Person Re-Identification (CVPR 2016 · similarity 0.88)
- Person Re-Identification by Camera Correlation Aware Feature Augmentation (TPAMI 2017 · similarity 0.87)
- Intra-Inter Domain Similarity for Unsupervised Person Re-Identification (TPAMI 2022 · similarity 0.87)
- Group Consistent Similarity Learning via Deep CRF for Person Re-Identification (CVPR 2018 · similarity 0.87)
- Deep Neural Networks with Inexact Matching for Person Re-Identification (NeurIPS 2016 · similarity 0.87)
- Cross-Domain Similarity Learning for Face Recognition in Unseen Domains (CVPR 2021 · similarity 0.87)
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.
