Attentive Task Interaction Network for Multi-Task Learning

TL;DR

The paper introduces ATI-Net, a novel multitask learning network that uses knowledge distillation and attention mechanisms to improve feature sharing and task performance, outperforming existing models with similar parameters.

Contribution

It presents a new approach combining knowledge distillation with attention in MTL, enhancing feature sharing and task performance.

Findings

Outperforms state-of-the-art MTL models like MTAN and PAD-Net

Uses knowledge distillation to improve contextualized feature sharing

Maintains comparable model complexity while enhancing accuracy

Abstract

Multitask learning (MTL) has recently gained a lot of popularity as a learning paradigm that can lead to improved per-task performance while also using fewer per-task model parameters compared to single task learning. One of the biggest challenges regarding MTL networks involves how to share features across tasks. To address this challenge, we propose the Attentive Task Interaction Network (ATI-Net). ATI-Net employs knowledge distillation of the latent features for each task, then combines the feature maps to provide improved contextualized information to the decoder. This novel approach to introducing knowledge distillation into an attention based multitask network outperforms state of the art MTL baselines such as the standalone MTAN and PAD-Net, with roughly the same number of model parameters.