Research map / Architectures and efficiency / RNNs and normalization
Controllable Dynamic Multi-Task Architectures
Dripta S. Raychaudhuri, Yumin Suh, Samuel Schulter, Xiang Yu, Masoud Faraki, Amit K. Roy-Chowdhury, Manmohan Chandraker
Check 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 dynamic tree architectures and preference-conditioned weights instead of a fixed shared backbone | Controllable dynamic multi-task network · NYU-v2 3-task | hypervolume: 11.72 -> 13.43 | Improved |
| Add cross-task adaptation with the weight hypernetwork (ablation) | Controllable dynamic multi-task network · NYU-v2 3-task | hypervolume: 12.42 -> 13.43 | Improved |
| Add cross-task adaptation with the weight hypernetwork (ablation) | Controllable dynamic multi-task network · NYU-v2 3-task | uniformity: 0.82 -> 0.76 | Hurt |
| Add cross-task adaptation with the weight hypernetwork (ablation) | Controllable dynamic multi-task network · CIFAR-100 | hypervolume: 0.003 -> 0.010 | Improved |
| Add the inactive branching loss (ablation) | Edge hypernetwork · NYU-v2 3-task | uniformity: 0.49 -> 0.76 | Improved |
| Add layer weighting to the active loss (ablation) | Edge hypernetwork · NYU-v2 3-task | uniformity: 0.51 -> 0.76 | Improved |
| Use task-affinity weighting in the active loss (ablation) | Edge hypernetwork · NYU-v2 3-task | uniformity: 0.51 -> 0.76 | Improved |
| Separate tasks into active and inactive groups (ablation) | Edge hypernetwork · NYU-v2 3-task | uniformity: 0.60 -> 0.76 | Improved |
| Increase cost preference c, which encourages more sharing and smaller architectures | Controllable dynamic multi-task network · NYU-v2 3-task | inference parameters: 41.06M -> 34.68M | Improved |
| Vary the Dirichlet parameter for sampling task preferences (ablation) | Controllable dynamic multi-task network · PASCAL-Context 3-task | hypervolume: 4.20; 4.26; 4.26; 4.26 | No clear effect |
Closest papers by meaning
- MTI-Net: Multi-Scale Task Interaction Networks for Multi-Task Learning (ECCV 2020 · similarity 0.88)
- Efficient Controllable Multi-Task Architectures (ICCV 2023 · similarity 0.88)
- Task Switching Network for Multi-Task Learning (ICCV 2021 · similarity 0.88)
- Task Adaptive Parameter Sharing for Multi-Task Learning (CVPR 2022 · similarity 0.88)
- Quantifying Task Priority for Multi-Task Optimization (CVPR 2024 · similarity 0.86)
- Selective Task Group Updates for Multi-Task Optimization (ICLR 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.
