Research map / Learning theory and algorithms / Prediction and decision losses
Iterative Projection and Matching: Finding Structure-Preserving Representatives and Its Application to Computer Vision
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 |
|---|---|---|---|
| Use IPM instead of uncertainty-based selection | UCF-101 active learning · UCF-101 split 1, early cycle (2 samples/class) | accuracy (%): 59.50 -> 64.93 | Helped |
| Combine IPM matching with uncertainty-based selection | UCF-101 active learning · UCF-101 split 1, 10 samples/class | accuracy (%): 77.18 -> 78.11 | Helped |
| Use IPM to choose representatives rather than random selection | UCF-101 classifier training · UCF-101, 1 representative/class | accuracy (%): 54.6 -> 65.3 | Helped |
| Select GAN training images with IPM rather than random selection | CR-GAN on Multi-PIE · Multi-PIE, 9 images/subject | identity dissimilarity: 0.5616 -> 0.553 | Helped |
| Use IPM rather than K-medoids for ImageNet representatives | ImageNet k-NN classification · ImageNet, 10 images/class | top-1 accuracy (%): 17.56 -> 25.26 | Helped |
| Use IPM rather than DS3 for UTE video summarization | UTE video summarization · UTE, ROUGE-SU | F-measure: 30.13 -> 31.53 | Helped |
| Use IPM rather than Submod-V+ | UTE video summarization · UTE, ROUGE-SU | F-measure: 34.15 -> 31.53 | Hurt |
| Use IPM rather than Submod-V+ | UTE video summarization · UTE, ROUGE-SU | Recall: 31.59 -> 29.09 | Hurt |
Closest papers by meaning
- IMP: Iterative Matching and Pose Estimation With Adaptive Pooling (CVPR 2023 · similarity 0.83)
- Joint Representative Selection and Feature Learning: A Semi-Supervised Approach (CVPR 2019 · similarity 0.83)
- Non-Uniform Subset Selection for Active Learning in Structured Data (CVPR 2017 · similarity 0.82)
- Search-Map-Search: A Frame Selection Paradigm for Action Recognition (CVPR 2023 · similarity 0.81)
- COMPOSE: Hypergraph Cover Optimization for Multi-view 3D Human Pose Estimation (NeurIPS 2026 · similarity 0.81)
- How to Use Diffusion Priors under Sparse Views? (NeurIPS 2024 · 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.