Process of the Dual-Channel Prototype Network (DCPN) Algorithm (IMAGE)

MedSight AI Research Lab

Process of the Dual-Channel Prototype Network (DCPN) Algorithm

Caption

Overview of the few-shot pathology image classification process constructed using the DCPN method. Initially, (A) the PVT model is pretrained based on self-supervised learning. Subsequently, (B) a dual-channel network is constructed in conjunction with a CNN to extract multi-scale features. Finally, (C) few-shot classification is realized using a similarity matrix based on multi-scale features and coupled with a soft voting strategy.

Credit

Hao Quan, Northeastern University

Usage Restrictions

News organizations may use or redistribute this image, with proper attribution.

License

Licensed content

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.