Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Label Distribution with Dirichlet Process Mixture Model
Minglong Wang, Weiwei Li, Yunan Lu, Xiuyi Jia
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 |
|---|---|---|---|
| use LDL-DPM instead of LDL-DPA | LDL-DPA baseline · SCUT-FBP | KL divergence: 1.1534 -> .3676 | Improved |
| use LDL-DPM instead of LDL-DPA | LDL-DPA baseline · M2B | KL divergence: .6749 -> .5302 | Improved |
| use LDL-DPM instead of LDL-DPA | LDL-DPA baseline · SBU 3DFE | KL divergence: .0550 -> .0510 | Improved |
| use LDL-DPM instead of LDL-DPA | LDL-DPA baseline · fbp5500 | KL divergence: .1045 -> .1059 | No clear effect |
| use LDL-DPM instead of LDL-DPA | LDL-DPA baseline · Movie | Cosine similarity: .9344 -> .9349 | No clear effect |
| use LDL-DPM instead of LDL-DPA | LDL-DPA baseline · M2B | Clark distance: 1.1714 -> 1.2214 | Hurt |
| use LDL-DPM instead of LDL-LRR | LDL-DPA baseline · Emotion6 | Intersection similarity: .5918 -> .5851 | Hurt |
| replace mixture with a single Dirichlet (add mixture) (ablation) | LDL-DPM · six LDL datasets | KL / Cosine | Improved |
| remove feature-conditioned gating (add gate) (ablation) | LDL-DPM · six LDL datasets | KL / Cosine | Improved |
| fix component number K=5 instead of adaptive (add DP-adaptive K) (ablation) | LDL-DPM · six LDL datasets | KL / Cosine | Improved |
Closest papers by meaning
- Generative Calibration of Inaccurate Annotation for Label Distribution Learning (AAAI 2024 · similarity 0.92)
- Adaptive-Grained Label Distribution Learning (AAAI 2025 · similarity 0.90)
- Approximately Correct Label Distribution Learning (ICML 2025 · similarity 0.89)
- Predicting Label Distribution from Ternary Labels (NeurIPS 2024 · similarity 0.89)
- Divide and Conquer: Learning Label Distribution with Subtasks (ICML 2025 · similarity 0.88)
- Concentration Distribution Learning from Label Distributions (ICML 2025 · similarity 0.88)
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.
