Martin Zurowietz

TL;DR: Biigle 2.0 as discussed by the authors annotates benthic fauna in marine image collections with tools customized to increase efficiency and effectiveness in the manual annotation process, and the software architecture of the system is described with different use-cases and future developments are discussed.

TL;DR: The MAIA approach represents a substantial improvement on the path to greater efficiency in the annotation of large benthic image collections and was able to annotate objects of interest with an average recall more than twice as fast as compared to “traditional” annotation methods.

TL;DR: In general, training only a classifier on top of convolutional layers pre-trained on extensive, but not domain-specific image data showed surprisingly high performance, indicating that domain adaptation to a novel taxonomic group can be feasible with a limited investment of effort.

TL;DR: This article presents Unsupervised Knowledge Transfer (UnKnoT), a new method to use the limited amount of training data more efficiently and employs a technique the authors call “scale transfer” and enhanced data augmentation to reuse existing training data for object detection of the same object classes in new image datasets.

TL;DR: In this article, eight image datasets were collected from a discrete area of polymetallic-nodule-rich seafloor by an AUV and several towed camera "ocean floor observation systems" deployed at various altitudes above the seaflores.