2019-05-24-lee19a.md

section	title	abstract	layout	series	id	month	tex_title	firstpage	lastpage	page	order	cycles	bibtex_author	author	date	address	publisher	container-title	volume	genre	issued	pdf	extras
Contributed Papers	DavinciGAN: Unpaired Surgical Instrument Translation for Data Augmentation	Recognizing surgical instruments in surgery videos is an essential process to describe surgeries, which can be used for surgery navigation and evaluation systems. In this paper, we argue that an imbalance problem is crucial when we train deep neural networks for recognizing surgical instruments using the training data collected from surgery videos since surgical instruments are not uniformly shown in a video. To address the problem, we use a generative adversarial network (GAN)-based approach to supplement insufficient training data. Using this approach, we could make training data have the balanced number of images for each class. However, conventional GANs such as CycleGAN and DiscoGAN, have a potential problem to be degraded in generating surgery images, and they are not effective to increase the accuracy of the surgical instrument recognition under our experimental settings. For this reason, we propose a novel GAN framework referred to as DavinciGAN, and we demonstrate that our method outperforms conventional GANs on the surgical instrument recognition task with generated training samples to complement the unbalanced distribution of human-labeled data.	inproceedings	Proceedings of Machine Learning Research	lee19a	0	DavinciGAN: Unpaired Surgical Instrument Translation for Data Augmentation	326	336	326-336	326	false	Lee, Kyungmoon and Choi, {Min-Kook} and Jung, Heechul	given family Kyungmoon Lee given family Min-Kook Choi given family Heechul Jung	2019-05-24		PMLR	Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning	102	inproceedings	date-parts 2019 5 24	http://proceedings.mlr.press/v102/lee19a/lee19a.pdf