Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
LaSO: Label-Set Operations Networks for Multi-Label Few-Shot Learning
Amit Alfassy, Leonid Karlinsky, Amit Aides, Joseph Shtok, Sivan Harary, Rogerio Feris, Raja Giryes, Alex M. Bronstein
This paper tried a common change: Add a reconstruction / auxiliary loss
Add symmetry reconstruction loss · MS-COCO 64 seen categories · mAP: 0.72 -> 0.76 Improved
What 987 accepted papers report for this change
- 882 improved
- 41 no clear effect
- 20 mixed
- 44 hurt
- OPTION: Optimal Transport–Guided Flow Matching for Incomplete and Unaligned Multi-View Clustering ICML 2026 · add reconstruction loss L_rec · Scene-15 (aligned, 70% missing) · ACC: 32.4 -> 37.5Improved
- Detached Skip-Links and $R$-Probe: Decoupling Feature Aggregation from Gradient Propagation for MLLM OCR ICML 2026 · add R-probe auxiliary loss · OCRBench · score: 714.0 -> 721.0Improved
- Rethinking Time-Series Imputation as Conditional Inference along Temporal Evolution ICML 2026 · add reconstruction loss on observed entries (joint vs missing-only supervision) · ETTh1 10% missing · Masked MAE: 0.2002 -> 0.1658Improved
See this change in detailMatched by the wording of each paper's own result, read by AI from the full text. Papers report gains more often than failures, so this shows what gets published, not the chance a change works.
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 |
|---|---|---|---|
| Use LaSO union augmentation instead of the strongest listed baseline, mixUP | MS-COCO 16 unseen categories, 1-shot classifier · MS-COCO 16 unseen categories, 1-shot | mAP: 40.2 -> 45.3 | Improved |
| Use LaSO union augmentation instead of the strongest listed baseline, mixUP | MS-COCO 16 unseen categories, 5-shot classifier · MS-COCO 16 unseen categories, 5-shot | mAP: 54.0 -> 58.1 | Improved |
| Use a learned LaSO operator instead of the analytic feature-wise expression | MS-COCO label-set subtraction · MS-COCO | mAP: 29.0 -> 43.0 | Improved |
| Use a learned LaSO operator instead of the analytic feature-wise expression | CelebA label-set subtraction · CelebA | mAP: 37.0 -> 69.0 | Improved |
| Use a learned operator instead of min(FX, FY) | MS-COCO label-set intersection · MS-COCO | mAP: 74.7 -> 77.0 | Improved |
| Use a learned operator instead of min(FX, FY) | CelebA label-set intersection · CelebA | mAP: 52.0 -> 48.0 | Hurt |
| Use a learned operator instead of max(FX, FY) | MS-COCO label-set union · MS-COCO | mAP: 76.5 -> 81.0 | Improved |
| Use a learned operator instead of max(FX, FY) | CelebA label-set union · CelebA | mAP: 47.0 -> 75 | Improved |
| Add symmetry reconstruction loss (ablation) | LaSO intersection · MS-COCO 64 seen categories | mAP: 0.72 -> 0.76 | Improved |
| Add mode-collapse reconstruction loss after adding symmetry loss (ablation) | LaSO subtraction · MS-COCO 64 seen categories | mAP: 0.39 -> 0.43 | Improved |
| Add both reconstruction losses (ablation) | LaSO operators · MS-COCO 64 seen categories | mAP: 3% − 9% | Improved |
| Add mode-collapse reconstruction loss during LaSO training (ablation) | LaSO retrieval outputs · MS-COCO seen and unseen categories | retrieved-sample diversity: 5%; 9% | Improved |
Closest papers by meaning
- Multi-Label Few-Shot Image Classification via Pairwise Feature Augmentation and Flexible Prompt Learning (AAAI 2025 · similarity 0.89)
- Inferring Prototypes for Multi-Label Few-Shot Image Classification with Word Vector Guided Attention (AAAI 2022 · similarity 0.88)
- Generative Multi-Label Zero-Shot Learning (TPAMI 2023 · similarity 0.87)
- Few-Shot Learning via Embedding Adaptation With Set-to-Set Functions (CVPR 2020 · similarity 0.87)
- Meta-Learning for Semi-Supervised Few-Shot Classification (ICLR 2018 · similarity 0.87)
- Simple Semantic-Aided Few-Shot Learning (CVPR 2024 · similarity 0.85)
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.
