Multitask learning with single gradient step update for task balancing

Sungjae Lee, Youngdoo Son

Research output: Contribution to journalArticlepeer-review

15 Scopus citations

Abstract

Multitask learning is a methodology to boost generalization performance and also reduce computational intensity and memory usage. However, learning multiple tasks simultaneously can be more difficult than learning a single task because it can cause imbalance among tasks. To address the imbalance problem, we propose an algorithm to balance between tasks at the gradient level by applying gradient-based meta-learning to multitask learning. The proposed method trains shared layers and task-specific layers separately so that the two layers with different roles in a multitask network can be fitted to their own purposes. In particular, the shared layer that contains informative knowledge shared among tasks is trained by employing single gradient step update and inner/outer loop training to mitigate the imbalance problem at the gradient level. We apply the proposed method to various multitask computer vision problems and achieve state-of-the-art performance.

Original languageEnglish
Pages (from-to)442-453
Number of pages12
JournalNeurocomputing
Volume467
DOIs
StatePublished - 7 Jan 2022

Keywords

  • Convolution neural network
  • Deep learning
  • Gradient-based meta learning
  • Multitask learning

Fingerprint

Dive into the research topics of 'Multitask learning with single gradient step update for task balancing'. Together they form a unique fingerprint.

Cite this