Auxiliary Learning by Implicit Differentiation
This addresses the problem of improving neural network performance through auxiliary tasks for researchers and practitioners in multi-task learning, though it appears incremental as it builds on existing implicit differentiation techniques.
The paper tackles the challenges of designing and combining auxiliary tasks in multi-task learning by proposing AuxiLearn, a framework using implicit differentiation to learn coherent loss combinations or generate novel auxiliary tasks, and finds it consistently outperforms competing methods in tasks like image segmentation and low-data learning.
Training neural networks with auxiliary tasks is a common practice for improving the performance on a main task of interest. Two main challenges arise in this multi-task learning setting: (i) designing useful auxiliary tasks; and (ii) combining auxiliary tasks into a single coherent loss. Here, we propose a novel framework, AuxiLearn, that targets both challenges based on implicit differentiation. First, when useful auxiliaries are known, we propose learning a network that combines all losses into a single coherent objective function. This network can learn non-linear interactions between tasks. Second, when no useful auxiliary task is known, we describe how to learn a network that generates a meaningful, novel auxiliary task. We evaluate AuxiLearn in a series of tasks and domains, including image segmentation and learning with attributes in the low data regime, and find that it consistently outperforms competing methods.