Enhanced task attention with adversarial learning for dynamic multi-task CNN

Yuchun Fang

doi:10.1016/j.patcog.2022.108672

2022-08-01

PRE

Yuchun Fang *

Menglu Zhou

张

张兆翔 (Zhaoxiang Zhang)

DOI：10.1016/j.patcog.2022.108672

原文链接

原文求助

摘要

En 中文

Multi-task deep learning is promising to solve multi-label multi-instance visual recognition tasks. How -ever, flexible information sharing in the task group might bring performance bottlenecks to an individual task. To tackle this problem, we propose a novel learning framework of multi-task Convolutional Neu-ral Network (CNN) to enhance task attention through conditionally tuning the Task Transfer Connections (TTC) with adversarial learning. For the dynamic multi-task CNN, we set up a shared subnet to extract shared features across multiple tasks and a task discriminator shared by all layers to distinguish features of all subnets. The adversarial training is introduced between the shared subnet and the task discrimi-nator to guide each task subnet to focus on its specific task. To apply adversarial learning to the com-plex labeling system of multiple tasks, we design an even-label strategy for the multi-task model with a shared subnet to make adversarial learning feasible for the complex labeling system of multiple tasks. As a result, the proposed model can constrain the shared subnet's learning unbiased to any single task and achieve task attention for all task subnets. Experimental results of the ablation study and the TTC analysis validate the effectiveness of the proposed approach.(c) 2022 Elsevier Ltd. All rights reserved.

Keyword:

Deep learning

Adversarial learning

Multi -task learning