Parallel Capsule Networks for Classification of White Blood Cells
Paper i proceeding, 2021

Capsule Networks (CapsNets) is a machine learning architecture proposed to overcome some of the shortcomings of convolutional neural networks (CNNs). However, CapsNets have mainly outperformed CNNs in datasets where images are small and/or the objects to identify have minimal background noise. In this work, we present a new architecture, parallel CapsNets, which exploits the concept of branching the network to isolate certain capsules, allowing each branch to identify different entities. We applied our concept to the two current types of CapsNet architectures, studying the performance for networks with different layers of capsules. We tested our design in a public, highly unbalanced dataset of acute myeloid leukaemia images (15 classes). Our experiments showed that conventional CapsNets show similar performance than our baseline CNN (ResNeXt-50) but depict instability problems. In contrast, parallel CapsNets can outperform ResNeXt-50, is more stable, and shows better rotational invariance than both, conventional CapsNets and ResNeXt-50.

Dynamic routing





Juan P. Vigueras-Guillén

AstraZeneca AB

Arijit Patra

AstraZeneca AB

Ola Engkvist

AstraZeneca AB

Chalmers, Data- och informationsteknik

Frank Seeliger

AstraZeneca AB

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

03029743 (ISSN) 16113349 (eISSN)

Vol. 12907 LNCS 743-752
9783030872335 (ISBN)

24th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2021
Virtual, Online, ,


Annan data- och informationsvetenskap





Mer information

Senast uppdaterat