Centerness-based Instance-aware Knowledge Distillation with Task-wise Mutual Lifting for Object Detection on Drone Imagery

TL;DR

This paper introduces a novel knowledge distillation approach tailored for drone imagery object detection, addressing challenges of small instances and complex backgrounds to improve accuracy efficiently.

Contribution

It proposes a task-wise Light-ML module and centerness-based distillation strategy specifically designed for drone imagery detection tasks.

Findings

Improves detection accuracy on drone datasets

Maintains computational efficiency comparable to existing methods

Enhances distillation effectiveness for small and complex instances

Abstract

Developing accurate and efficient detectors for drone imagery is challenging due to the inherent complexity of aerial scenes. While some existing methods aim to achieve high accuracy by utilizing larger models, their computational cost is prohibitive for drones. Recently, Knowledge Distillation (KD) has shown promising potential for maintaining satisfactory accuracy while significantly compressing models in general object detection. Considering the advantages of KD, this paper presents the first attempt to adapt it to object detection on drone imagery and addresses two intrinsic issues: (1) low foreground-background ratio and (2) small instances and complex backgrounds, which lead to inadequate training, resulting insufficient distillation. Therefore, we propose a task-wise Lightweight Mutual Lifting (Light-ML) module with a Centerness-based Instance-aware Distillation (CID) strategy. The Light-ML module mutually harmonizes the classification and localization branches by channel shuffling and convolution, integrating teacher supervision across different tasks during back-propagation, thus facilitating training the student model. The CID strategy extracts valuable regions surrounding instances through the centerness of proposals, enhancing distillation efficacy. Experiments on the VisDrone, UAVDT, and COCO benchmarks demonstrate that the proposed approach promotes the accuracies of existing state-of-the-art KD methods with comparable computational requirements. Codes will be available upon acceptance.