Affiliation:
1. Department of Automatic Control and Robotics , AGH University of Science and Technology , Krakow , Poland
2. SOLARIS National Synchrotron Radiation Centre , Jagiellonian University , Krakow , Poland
Abstract
Abstract
Deep learning methods, used in machine vision challenges, often face the problem of the amount and quality of data. To address this issue, we investigate the transfer learning method. In this study, we briefly describe the idea and introduce two main strategies of transfer learning. We also present the widely-used neural network models, that in recent years performed best in ImageNet classification challenges. Furthermore, we shortly describe three different experiments from computer vision field, that confirm the developed algorithms ability to classify images with overall accuracy 87.2-95%. Achieved numbers are state-of-the-art results in melanoma thickness prediction, anomaly detection and Clostridium di cile cytotoxicity classification problems.
Reference17 articles.
1. [1] Bowyer K. W., Chawla N. V., Hall L. O., and Kegelmeyer W. P. SMOTE: synthetic minority over-sampling technique. CoRR, abs/1106.1813, 2011.
2. [2] Chollet F. Xception: Deep learning with depthwise separable convolutions. CoRR, abs/1610.02357, 2016.10.1109/CVPR.2017.195
3. [3] Garland M., Jaworek-Korjakowska J., Libal U., Bogyo M., and M. S. An automatic analysis system for high-throughput clostridium di cile toxin activity screening. Applied Science, 8(1512), 2018.10.3390/app8091512
4. [4] He K., Zhang X., Ren S., and Sun J. Deep residual learning for image recognition. CoRR, abs/1512.03385, 2015.
5. [5] Huang G., Liu Z., van der Maaten L., and Weinberger K. Q. Densely connected convolutional networks, 2016.10.1109/CVPR.2017.243
Cited by
22 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献