Abstract | ||
---|---|---|
Multi-task learning commonly encounters competition for resources among tasks, specifically when model capac-ity is limited. This challenge motivates models which al-low control over the relative importance of tasks and total compute cost during inference time. In this work, we pro-pose such a controllable multi-task network that dynami-cally adjusts its architecture and weights to match the de-sired task preference as well as the resource constraints. In contrast to the existing dynamic multi-task approaches that adjust only the weights within a fixed architecture, our approach affords the flexibility to dynamically control the total computational cost and match the user-preferred task importance better. We propose a disentangled training of two hype rnetwo rks, by exploiting task affinity and a novel branching regularized loss, to take input prefer-ences and accordingly predict tree-structured models with adapted weights. Experiments on three multi-task bench-marks, namely PASCAL-Context, NYU-v2, and CIFAR-100, show the efficacy of our approach. Project page is available at https://www.nec-labs.com/-mas/DYMU. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1109/CVPR52688.2022.01068 | IEEE Conference on Computer Vision and Pattern Recognition |
Keywords | DocType | Volume |
Deep learning architectures and techniques, Vision applications and systems | Conference | 2022 |
Issue | Citations | PageRank |
1 | 0 | 0.34 |
References | Authors | |
0 | 7 |
Name | Order | Citations | PageRank |
---|---|---|---|
Dripta S. Raychaudhuri | 1 | 0 | 0.34 |
Yumin Suh | 2 | 49 | 4.38 |
Samuel Schulter | 3 | 158 | 11.58 |
Xiang Yu | 4 | 29 | 5.16 |
Masoud Faraki | 5 | 0 | 1.35 |
Amit K. Roy Chowdhury | 6 | 1153 | 73.96 |
Manmohan Chandraker | 7 | 250 | 21.13 |