Jonas Prellberg

TL;DR: Evaluating tool annotation algorithms based on deep learning for cataract surgery finds that the quality of their annotations are compared to that of human interpretations, and it is expected that they will guide the design of efficient surgery monitoring tools in the near future.

TL;DR: This work presents a simple, yet effective classification approach using a ResNeXt convolutional neural network with Squeeze-and-Excitation modules for leukemia classification, evaluated in the C-NMC online challenge.

TL;DR: It is shown that an evolutionary algorithm saves training time during the network architecture optimization, if learned network weights are inherited over generations by Lamarckian evolution.

TL;DR: This work implements an evolutionary algorithm that executes entirely on the GPU, which allows to efficiently batch-evaluate a whole population of networks and explores crossover operators and finds that unprincipled random uniform crossover performs extremely well.

TL;DR: In this paper, a system that can distinguish 21 different tools in cataract surgery videos is created using the successful residual network architecture, which is used for real-time computer assistance for the surgeon.