Research map / Label-efficient and robust learning / Domain adaptation and generalization
Conformalized Hierarchical Calibration for Uncertainty-Aware Adaptive Hashing
Junyu Luo, Jinsheng Huang, Yang Xu, Lutong Zou, Xiao Luo, Bohan Wu, Yifan Wang, Wei Ju, Ming Zhang
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 semantic-level conformal calibration (ablation) | COLA · Office-Home (64-bit) | avg mAP: 54.53 -> 57.31 | Improved |
| add representation-level bit calibration (ablation) | COLA · Office-Home (64-bit) | avg mAP: 55.21 -> 57.31 | Improved |
| add self-regulating mechanism (ablation) | COLA · Office-Home (64-bit) | avg mAP: 54.74 -> 57.31 | Improved |
| add semantic calibration alone to baseline (ablation) | COLA (None) · Office-Home (64-bit) | avg mAP: 49.14 -> 49.43 | No clear effect |
| add representation calibration alone to baseline (ablation) | COLA (None) · Office-Home (64-bit) | avg mAP: 49.14 -> 51.76 | Improved |
| add self-regulating alone to baseline (ablation) | COLA (None) · Office-Home (64-bit) | avg mAP: 49.14 -> 51.55 | Improved |
| use target-centroid calibration set instead of random sampling (ablation) | COLA · Office-Home Ar→Re | mAP: 55.21 -> 56.34 | Improved |
| use target-centroid calibration set instead of density-aware sampling (ablation) | COLA · Office-Home Ar→Re | mAP: 55.85 -> 56.34 | Improved |
| use dynamic alpha instead of fixed alpha (ablation) | COLA · Office-Home | mAP: 55.22 -> 57.31 | Improved |
| use dynamic alpha instead of fixed alpha (ablation) | COLA · Digits | mAP: 69.57 -> 70.41 | Improved |
| use EMA smoothing 0.7 instead of 0.9 (ablation) | COLA · Office-Home | mAP: 56.93 -> 57.31 | Improved |
| use UWHD instead of standard Hamming distance (ablation) | COLA · Office-Home (64-bit) | avg mAP: 56.23 -> 57.31 | Improved |
Closest papers by meaning
- Robust Domain Adaptive Hashing via Structural Noise Modeling and Correction (AAAI 2026 · similarity 0.93)
- Domain Adaptive Hashing Retrieval via VLM Assisted Pseudo-Labeling and Dual Space Adaptation (NeurIPS 2025 · similarity 0.90)
- Effective Comparative Prototype Hashing for Unsupervised Domain Adaptation (AAAI 2024 · similarity 0.90)
- IDEA: An Invariant Perspective for Efficient Domain Adaptive Image Retrieval (NeurIPS 2023 · similarity 0.90)
- KALAHash: Knowledge-Anchored Low-Resource Adaptation for Deep Hashing (AAAI 2025 · similarity 0.88)
- Deep Evidential Hashing for Trustworthy Cross-Modal Retrieval (AAAI 2025 · similarity 0.86)
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.
