Research map / Label-efficient and robust learning / Recognition and long-tail
Efficient Multiple Instance Metric Learning Using Weakly Supervised Data
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| jointly learn the metric and instance assignments from bag-level labels | MIMLCA · Labeled Yahoo! News, setting (b) | accuracy: 56.1 ± 3.3 -> 65.3 ± 3.7 | Helped |
| use automatically extracted bag labels with joint metric and assignment learning | MIMLCA · Labeled Yahoo! News, setting (c) | accuracy: 52.6 ± 13.0 -> 63.2 ± 4.7 | Helped |
| use MIMLCA instead of MildML | weakly supervised face identification · Labeled Yahoo! News, setting (b) | training time (seconds): 7,352 -> 163 | Helped |
| use MIMLCA instead of MIML [12] | automated image annotation · Corel5K | Average Precision: 0.535 -> 0.575 | Helped |
| use MIMLCA instead of MIML [12] | automated image annotation · Corel5K | training time (seconds): 884 seconds -> 24 seconds | Helped |
| replace the linear model with the kRBF χ² nonlinear model (ablation) | MIMLCA · Labeled Yahoo! News, setting (b), α = 0.25 | accuracy: 81.9 ± 2.5 -> 88.5 ± 2.1 | Helped |
| vary α from 0 to 0.5 (ablation) | Linear MIMLCA · Labeled Yahoo! News, setting (b) | accuracy: 74.2 ± 2.7 -> 87.7 ± 1.9 | Helped |
| increase α from 0.25 to 1.2 (ablation) | kRBF χ² MIMLCA · Labeled Yahoo! News, setting (b) | accuracy: 88.5 ± 2.1 -> 78.6 ± 3.6 | Hurt |
| increase α from 0.25 to 1.2 (ablation) | kRBF χ² MIMLCA · Labeled Yahoo! News, setting (c) | accuracy: 88.5 ± 2.1 -> 78.6 ± 3.6 | Hurt |
Closest papers by meaning
- Self-Taught Metric Learning Without Labels (CVPR 2022 · similarity 0.86)
- Weakly Supervised Instance Segmentation using the Bounding Box Tightness Prior (NeurIPS 2019 · similarity 0.86)
- Divide and Conquer the Embedding Space for Metric Learning (CVPR 2019 · similarity 0.85)
- Multiple Instance Detection Network With Online Instance Classifier Refinement (CVPR 2017 · similarity 0.85)
- T-MDML: Triplet-based Multiple Distance Metric Learning for Multi-Instance Multi-Label Classification with Label Correlation (AAAI 2025 · similarity 0.84)
- WISH: Weakly Supervised Instance Segmentation using Heterogeneous Labels (CVPR 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.