The success of capsule networks lies in their ability to preserve more information about the input by replacing max-pooling layers with convolutional strides and dynamic routing, allowing for preservation of part-whole relationships in the data.
Research in Biomedical Imaging Analysis
SegCaps: Capsule for object segmentation
The proposed convolutional-deconvolutional capsule network, called SegCaps, shows strong results for the task of object segmentation with substantial decrease in parameter space.
XCaps: Explainable Capsules
Our explainable capsule network, X-Caps, encodes high-level visual object attributes within the vectors of its capsules, then forms predictions based solely on these human-interpretable features.
DCaps: Diagnosing Colorectal Polyps in the Wild with Capsule Networks
In this study, we design a novel capsule network architecture (D-Caps) to improve the viability of optical biopsy of colorectal polyps. Our proposed method introduces several technical novelties including a novel capsule architecture with a capsule-average pooling (CAP) method to improve efficiency in large-scale image classification
1. Rodney Lalonde, Naji Khosravan, Ulas Bagci, Deformable Capsules for Object Detection, U.S. Patent No. 11,514,579. 29 Nov. 2022.
2. Harish Ravi Prakash; Syed Muhammad Anwar; Ulas Bagci, Variational Capsule Encoder, ICPR, 2020.
3. Merey Sarsengeldin; Sanim Imatayeva; Nurmukhamed Abeuov; Myrzakhan Naukhanov; Abdullah Said Erdogan; Debesh Jha; Ulas Bagci, Gastrointestinal Disease Diagnosis with Hybrid Model of Capsules and CNNs, 2023 IEEE International Conference on Electro Information Technology, 2023