Research map / Label-efficient and robust learning / Recognition and long-tail
Combining Task Predictors via Enhancing Joint Predictability
Kwang In Kim, Christian Richardt, Hyung Jin Chang
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 joint denoising of references along with the target (ablation) | NPC · OSR attribute ranking | accuracy improvement over f0: 1.96 -> 2.13 | Improved |
| add automatic relevance determination (ablation) | NPC · OSR attribute ranking | accuracy improvement over f0: 1.88 -> 2.13 | Improved |
| use nonlinear instead of linear combination (XOR target) | LPC to NPC · Toy problem 2 | accuracy: 74.24 -> 100 | Improved |
| use joint (linear) relevance instead of pairwise | OPC to LPC · Toy problem 1 | accuracy: 67.24 -> 100 | Improved |
| use NPC instead of OPC | predictor combination · Shoes classification | accuracy (%): 58.52 -> 62.87 | Improved |
| use NPC instead of OPC | predictor combination · Pubfig classification | accuracy (%): 82.55 -> 86.51 | Improved |
| use NPC instead of OPC | predictor combination · OSR classification | accuracy (%): 77.16 -> 79.71 | Improved |
| use NPC instead of OPC | predictor combination · aPascal classification | accuracy (%): 39.77 -> 40.34 | Improved |
| use NPC instead of OPC | predictor combination · CUB classification | accuracy (%): 67.75 -> 68.26 | Improved |
| raise number of basis vectors N' from 300 to 1,000 (ablation) | NPC · low-rank GP approximation | accuracy | No clear effect |
| use NPC on AWA2 versus OPC and MTL | predictor combination · AWA2 ranking (80 attributes) | attributes with significant gain: 74 vs 8 and 33 | Improved |
| use NPC on Zap50K | NPC · Zap50K ranking | significant gains: 0 significantly better, 4 on par | No clear effect |
Closest papers by meaning
- Identifying and Mitigating Spurious Correlation in Multi-Task Learning (CVPR 2025 · similarity 0.84)
- Towards Principled Task Grouping for Multi-Task Learning (NeurIPS 2026 · similarity 0.83)
- No More Tuning: Prioritized Multi-Task Learning with Lagrangian Differential Multiplier Methods (AAAI 2025 · similarity 0.82)
- Multi-Task Dense Prediction via Mixture of Low-Rank Experts (CVPR 2024 · similarity 0.82)
- AdaMerging: Adaptive Model Merging for Multi-Task Learning (ICLR 2024 · similarity 0.82)
- In-Context Learning Strategies Emerge Rationally (NeurIPS 2025 · 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.
