Controllable Dynamic Multi-Task Architectures

Dripta S. Raychaudhuri, Yumin Suh, Samuel Schulter, Xiang Yu, Masoud Faraki, Amit K. Roy-Chowdhury, Manmohan Chandraker; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 10955-10964


Multi-task learning commonly encounters competition for resources among tasks, specifically when model capacity is limited. This challenge motivates models which allow control over the relative importance of tasks and total compute cost during inference time. In this work, we propose such a controllable multi-task network that dynamically adjusts its architecture and weights to match the desired 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 hypernetworks, by exploiting task affinity and a novel branching regularized loss, to take input preferences and accordingly predict tree-structured models with adapted weights. Experiments on three multi-task benchmarks, namely PASCAL-Context, NYU-v2, and CIFAR-100, show the efficacy of our approach. Project page is available at mas/DYMU.

Related Material

[pdf] [supp] [arXiv]
@InProceedings{Raychaudhuri_2022_CVPR, author = {Raychaudhuri, Dripta S. and Suh, Yumin and Schulter, Samuel and Yu, Xiang and Faraki, Masoud and Roy-Chowdhury, Amit K. and Chandraker, Manmohan}, title = {Controllable Dynamic Multi-Task Architectures}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)}, month = {June}, year = {2022}, pages = {10955-10964} }