Research map / Label-efficient and robust learning / Recognition and long-tail
Dense Interspecies Face Embedding
Sejong Yang, Subin Jeon, Seonghyeon Nam, Seon Joo Kim
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 StyleGAN2 distillation loss LK2 to CSE distillation LK1 (ablation) | DIFE · Human+Dog, pixel error | pixel error: 19.33 -> 13.59 | Improved |
| add semantic matching losses LM1 and LM2 on top of LK1+LK2 (ablation) | DIFE · Human+Dog, pixel error | pixel error: 13.59 -> 11.73 | Improved |
| add CSE distillation loss LK1 (ablation) | DIFE · Human+Dog, pixel error | pixel error: 17.45 -> 11.73 | Improved |
| add one semantic matching loss (the other omitted) (ablation) | DIFE · Human+Dog, pixel error | pixel error: 13.03 -> 11.73 | Improved |
| add the other semantic matching loss (ablation) | DIFE · Human+Dog, pixel error | pixel error: 12.22 -> 11.73 | Improved |
| use StyleGAN2 layer 7 instead of layer 5 (ablation) | DIFE · Human+Dog, pixel error | pixel error: 13.80 -> 11.73 | Improved |
| use StyleGAN2 layer 7 instead of layer 9 (ablation) | DIFE · Human+Dog, pixel error | pixel error: 14.78 -> 11.73 | Improved |
| DIFE instead of CATs on interspecies transfer | CATs · MAFL/AP-10K Human+Dog | pixel error: 14.78 -> 11.73 | Improved |
| DIFE instead of DVE on interspecies transfer | DVE · MAFL/AP-10K Dog+Cat | pixel error: 6.75 -> 6.51 | Improved |
| DIFE instead of CATs | CATs · MAFL/AP-10K Human+Cat | pixel error: 17.63 -> 11.00 | Improved |
| DIFE instead of CATs on denser landmarks | CATs · WFLW/AnimalWeb Human+Wild | pixel error: 22.20 -> 14.03 | Improved |
| DIFE instead of DVE on denser landmarks | DVE · WFLW/AnimalWeb Human+Cat | pixel error: 13.28 -> 12.84 | Improved |
Closest papers by meaning
- Deep Cross-species Feature Learning for Animal Face Recognition via Residual Interspecies Equivariant Network (ECCV 2020 · similarity 0.87)
- Interspecies Knowledge Transfer for Facial Keypoint Detection (CVPR 2017 · similarity 0.87)
- Learning the 3D Fauna of the Web (CVPR 2024 · similarity 0.82)
- Reconstructing Animals and the Wild (CVPR 2025 · similarity 0.82)
- SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild (NeurIPS 2026 · similarity 0.82)
- Generating the Wild: Individual-Consistent Image-to-Video Generation for Wildlife (NeurIPS 2026 · similarity 0.81)
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.
