Research map / Label-efficient and robust learning / Recognition and long-tail
DPF: Learning Dense Prediction Fields With Weak Supervision
Xiaoxue Chen, Yuhang Zheng, Yupeng Zheng, Qiang Zhou, Hao Zhao, Guyue Zhou, Ya-Qin Zhang
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 DPF to the DPT baseline | DPT · PASCALContext | mIoU: 40.4 -> 45.3 | Improved |
| add DPF to the DPT baseline | DPT · ADE20K | mIoU: 30.4 -> 33.8 | Improved |
| add DPF to the DPT baseline | DPT · IIW | WHDR: 14.0 -> 11.9 | Improved |
| add auxiliary supervision to the dense prediction backbone (ablation) | DPF · PASCAL | mIoU: 40.0 (-5.3) -> 45.3 | Improved |
| add auxiliary supervision to the dense prediction backbone (ablation) | DPF · ADE20K | mIoU: 28.3 (-5.5) -> 33.8 | Improved |
| add auxiliary supervision to the dense prediction backbone (ablation) | DPF · IIW | WHDR: 25.2 (-13.3) -> 11.9 | Improved |
| add the guidance encoder and guidance features (ablation) | DPF · PASCAL | mIoU: 44.5 (-0.8) -> 45.3 | Improved |
| add the guidance encoder and guidance features (ablation) | DPF · ADE20K | mIoU: 32.9 (-0.9) -> 33.8 | Improved |
| add the guidance encoder and guidance features (ablation) | DPF · IIW | WHDR: 12.5 (-0.6) -> 11.9 | Improved |
| use learned interpolation weights instead of relative-distance weights (ablation) | DPF · PASCAL | mIoU: 44.2 -> 45.3 | Improved |
| use learned interpolation weights instead of relative-distance weights (ablation) | DPF · ADE20K | mIoU: 32.1 -> 33.8 | Improved |
| use learned interpolation weights instead of relative-distance weights (ablation) | DPF · IIW | WHDR: 12.9 -> 11.9 | Improved |
Closest papers by meaning
- A General Framework for Learning from Weak Supervision (ICML 2024 · similarity 0.82)
- Towards Safe Weakly Supervised Learning (TPAMI 2019 · similarity 0.81)
- Learning SMaLL Predictors (NeurIPS 2018 · similarity 0.81)
- Gaussian Process Density Counting from Weak Supervision (ECCV 2016 · similarity 0.81)
- Accelerating Neural Field Training via Soft Mining (CVPR 2024 · similarity 0.80)
- Deep Probabilistic Supervision for Image Classification (NeurIPS 2026 · similarity 0.79)
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.
