Leveraging knowledge distillation for partial multi-task learning from multiple remote sensing datasets

TL;DR

This paper introduces a knowledge distillation approach to improve partial multi-task learning in remote sensing, enabling effective training with datasets lacking full annotations for all tasks.

Contribution

It proposes a novel method using knowledge distillation to enhance partial multi-task learning without requiring complete ground truth annotations.

Findings

Improved performance on semantic segmentation and object detection tasks.

Effective handling of datasets with partial annotations.

Validated on ISPRS 2D Semantic Labeling dataset.

Abstract

Partial multi-task learning where training examples are annotated for one of the target tasks is a promising idea in remote sensing as it allows combining datasets annotated for different tasks and predicting more tasks with fewer network parameters. The na\"ive approach to partial multi-task learning is sub-optimal due to the lack of all-task annotations for learning joint representations. This paper proposes using knowledge distillation to replace the need of ground truths for the alternate task and enhance the performance of such approach. Experiments conducted on the public ISPRS 2D Semantic Labeling Contest dataset show the effectiveness of the proposed idea on partial multi-task learning for semantic tasks including object detection and semantic segmentation in aerial images.