Research map / Label-efficient and robust learning / Recognition and long-tail
Few-Shot Partial Multi-View Learning
Yuan Zhou, Yanrong Guo, Shijie Hao, Richang Hong, Jiebo Luo
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 UGDA on top of PTMAP | PTMAP · Cub-googlenet-doc2vec 3-way 1-shot, view-missing rate 0.5 | accuracy (%): 56.87 -> 77.52 | Improved |
| add UGDA on top of PTMAP | PTMAP · Caltech102 5-way 1-shot, view-missing rate 0.5 | accuracy (%): 30.99 -> 40.76 | Improved |
| add UGDA on top of PTMAP with no missing views | PTMAP · Caltech102 5-way 1-shot, view-missing rate 0 | accuracy (%): 59.58 -> 56.58 | Hurt |
| add UGDA on top of RFS | RFS · Animal 5-way 1-shot, view-missing rate 0.5 | accuracy (%): 48.21 -> 63.32 | Improved |
| add UGDA on top of RFS | RFS · ORL 5-way 1-shot, view-missing rate 0 | accuracy (%): 86.51 -> 90.95 | Improved |
| use UGDA instead of the strongest few-shot baseline at eta=0.5 | PTMAP · Scene15 3-way 1-shot | accuracy (%): 50.68 -> 62.17 | Improved |
| use UGDA instead of PTMAP with complete views | PTMAP · Handwritten 3-way 1-shot, view-missing rate 0 | accuracy (%): 89.14 -> 89.08 | No clear effect |
| use UGDA instead of COMPLETER at high missing rate | COMPLETER · Scene15 3-way 1-shot, view-missing rate 0.5 | accuracy (%): 43.20 -> 62.17 | Improved |
| use UGDA instead of ICMSC at high missing rate | ICMSC · ORL 5-way 1-shot, view-missing rate 0.5 | accuracy (%): 20.02 -> 64.78 | Improved |
| increase view-missing rate from 0 to 0.5 for UGDA | UGDA · Cub-googlenet-doc2vec 3-way 1-shot | accuracy (%): 95.59 -> 76.69 | Hurt |
Closest papers by meaning
- Dense Classification and Implanting for Few-Shot Learning (CVPR 2019 · similarity 0.86)
- Partial Is Better Than All: Revisiting Fine-tuning Strategy for Few-shot Learning (AAAI 2021 · similarity 0.85)
- Dataset Bias in Few-Shot Image Recognition (TPAMI 2022 · similarity 0.85)
- Rethinking Generalization in Few-Shot Classification (NeurIPS 2022 · similarity 0.85)
- Pushing the Limit of Fine-Tuning for Few-Shot Learning: Where Feature Reusing Meets Cross-Scale Attention (AAAI 2024 · similarity 0.85)
- Biased Incomplete Multi-View Learning (AAAI 2025 · 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.
