Quantization-based deep diversified ensemble for medical image segmentation

Jiawei Zhang

doi:10.1016/j.engappai.2025.112242

Quantization-based deep diversified ensemble for medical image segmentation

2025-10-08

0

PRE

AI

张彦春 (Yanchun Zhang) *

DOI：10.1016/j.engappai.2025.112242

Original

Original request for help

Save

Abstract

En 中文

• We provide valuable insights into the importance of the diversity of the internal feature from ensemble learners. • We propose a deep diversified loss to enhance diversity by directly diversifying the internal features from different base learners. • We propose a deep diversified quantization to enhance diversity that selectively preserves the randomness brought about by quantization. • We propose a deep ensemble, in which a meta learner is used to fuse the internal diversified features from various base learners. • Comprehensive experimental results across five popular datasets and various networks demonstrate the superiority of our method over existing approaches.

Quantization-based deep diversified ensemble for medical image segmentation

Abstract

Journal

Organization