MEDIA ANALYTICS
PROJECTS
PEOPLE
PUBLICATIONS
PATENTS
Multi-task learning commonly encounters competition for resources among tasks when model capacity is limited. We develop neural architectures that allow control over the relative importance of tasks and total compute cost during inference time. Our controllable multi-task networks dynamically adjust architecture and weights to match desired task preferences as well as resource constraints. Our dynamic networks allow optimizing performance for continually changing varying user needs, without incurring the heavy computational overhead to train and save models for various scenarios.
Team Members: Abhishek Aich
Publication Tags (project tag): dymu
We aim to train a multi-task model such that users can adjust the desired compute budget and relative importance of task performances after deployment, without retraining. This enables optimizing performance for dynamically varying user needs, without heavy computational overhead to train and save models
This paper presents an approach to train a unified deep network that simultaneously solves multiple human-related tasks. A multi-task framework is favorable for sharing information across tasks under restricted computational resources. However, tasks not only share information but may also compete for
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