Research map / Image generation / Image restoration
Beyond IID: Learning to Combine Non-IID Metrics for Vision Tasks
Yinghuan Shi, Wenbin Li, Yang Gao, Longbing Cao, Dinggang Shen
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 |
|---|---|---|---|
| replace IID metric learning (LMCA) with non-IID metric learning (NIME-CK) | metric learning · MSRC segmentation | pixel error rate (%): 5.03 -> 3.27 | Improved |
| replace IID metric learning (LMNN) with non-IID metric learning (NIME-CK) | metric learning · MSRC segmentation | pixel error rate (%): 4.83 -> 3.27 | Improved |
| combine three non-IID representations (NIME-CK) instead of the best single one (NIME-MP) (ablation) | NIME · MSRC segmentation | pixel error rate (%): 3.31 -> 3.27 | Improved |
| replace LMNN with NIME-CK non-IID metric | histology image identification · Histology (hand-crafted) | accuracy (%): 80.0 -> 89.0 | Improved |
| replace LMNN with NIME-CK non-IID metric | histology image identification · Histology (deep features) | AUC (%): 86.6 -> 96.9 | Improved |
| combine non-IID representations (NIME-CK) instead of single direct-product (NIME-DP) (ablation) | NIME · Histology (hand-crafted) | accuracy (%): 86.0 -> 89.0 | Improved |
| combine non-IID representations (NIME-CK) instead of single NIME-MP (ablation) | NIME · Histology (deep features) | specificity (%): NIME-DP 89.6 vs NIME-CK 88.5 | Mixed |
| switch from hand-crafted to deeply-learned (LeNet) cell features | NIME-CK · Histology identification | accuracy (%): 89.0 -> 90.0 | Improved |
| vary number of spatial neighbors from 3 to 7 (ablation) | NIME · Histology identification | accuracy: HC: 0.84, 0.89, 0.89, 0.87, 0.87; DL: 0.89, 0.91, 0.90, 0.89, 0.88 | Mixed |
Closest papers by meaning
- Metric Convolutions: A Unifying Theory to Adaptive Image Convolutions (ICCV 2025 · similarity 0.83)
- Unsupervised Deep Metric Learning with Transformed Attention Consistency and Contrastive Clustering Loss (ECCV 2020 · similarity 0.83)
- Deep Metric Learning with Graph Consistency (AAAI 2021 · similarity 0.82)
- Energy Confused Adversarial Metric Learning for Zero-Shot Image Retrieval and Clustering (AAAI 2019 · similarity 0.82)
- MetricOpt: Learning To Optimize Black-Box Evaluation Metrics (CVPR 2021 · similarity 0.82)
- MIC: Mining Interclass Characteristics for Improved Metric Learning (ICCV 2019 · similarity 0.82)
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.
