TY - GEN
T1 - Fine-tuning convolutional neural networks for biomedical image analysis
T2 - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017
AU - Zhou, Zongwei
AU - Shin, Jae
AU - Zhang, Lei
AU - Gurudu, Suryakanth
AU - Gotway, Michael
AU - Liang, Jianming
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/11/6
Y1 - 2017/11/6
N2 - Intense interest in applying convolutional neural networks (CNNs) in biomedical image analysis is wide spread, but its success is impeded by the lack of large annotated datasets in biomedical imaging. Annotating biomedical images is not only tedious and time consuming, but also demanding of costly, specialty-oriented knowledge and skills, which are not easily accessible. To dramatically reduce annotation cost, this paper presents a novel method called AIFT (active, incremental fine-tuning) to naturally integrate active learning and transfer learning into a single framework. AIFT starts directly with a pre-trained CNN to seek "worthy" samples from the unannotated for annotation, and the (fine-tuned) CNN is further fine-tuned continuously by incorporating newly annotated samples in each iteration to enhance the CNN's performance incrementally. We have evaluated our method in three different biomedical imaging applications, demonstrating that the cost of annotation can be cut by at least half. This performance is attributed to the several advantages derived from the advanced active and incremental capability of our AIFT method.
AB - Intense interest in applying convolutional neural networks (CNNs) in biomedical image analysis is wide spread, but its success is impeded by the lack of large annotated datasets in biomedical imaging. Annotating biomedical images is not only tedious and time consuming, but also demanding of costly, specialty-oriented knowledge and skills, which are not easily accessible. To dramatically reduce annotation cost, this paper presents a novel method called AIFT (active, incremental fine-tuning) to naturally integrate active learning and transfer learning into a single framework. AIFT starts directly with a pre-trained CNN to seek "worthy" samples from the unannotated for annotation, and the (fine-tuned) CNN is further fine-tuned continuously by incorporating newly annotated samples in each iteration to enhance the CNN's performance incrementally. We have evaluated our method in three different biomedical imaging applications, demonstrating that the cost of annotation can be cut by at least half. This performance is attributed to the several advantages derived from the advanced active and incremental capability of our AIFT method.
UR - http://www.scopus.com/inward/record.url?scp=85040371714&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85040371714&partnerID=8YFLogxK
U2 - 10.1109/CVPR.2017.506
DO - 10.1109/CVPR.2017.506
M3 - Conference contribution
AN - SCOPUS:85040371714
T3 - Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017
SP - 4761
EP - 4772
BT - Proceedings - 30th IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 21 July 2017 through 26 July 2017
ER -