EfficientDet with Knowledge Distillation and Instance Whitening for Real-Time and Generalisable Polyp Detection

Raneem Toman, Venkataraman Subramanian, Sharib Ali

Abstract: Despite numerous techniques developed for polyp detection, the issue of generalisability to new centres and populations while maintaining fast inference persists. To address this, we compile a multicentre train set consisting of 1941 images, and use it to push the generalisability of the real-time EfficientDet for polyp detection by employing a knowledge distillation teacher-student architecture and instance whitening. We train a large EfficientDet teacher on the combined Kvasir-Seg and PolypGen datasets and subsequently use its softened output probabilities to guide a smaller EfficientDet student, aiming to improve generalisability and speed while preserving performance. To further enhance generalisability across different clinical settings and patient populations, we integrate instance whitening into the backbones of both teacher and student networks to mitigate domain-specific variations, especially in patient populations like IBD. We compare our model with the state-of-the-art (SOTA) models DETR, RetinaNet, and Faster R-CNN under detection and inference metrics on the following benchmark datasets: Kvasir-SEG (seen centre and population), PolypGen-C6 (unseen centre), and our in-house IBD dataset (unseen centre and population). Our approach improves AP50:95 over the best performing benchmark -RetinaNet- by 3.97%, 10.67%, and 11.07% on the Kvasir-SEG, PolypGen-C6, and IBD datasets respectively, highlighting the significant role of instance whitening and distillation in boosting the model’s ability to generalise to new, unseen data distributions, particularly in the in-house IBD dataset, making our model clinically applicable.

External IDs:doi:10.1007/978-3-031-98691-8_23