| Title | A convolutional generative adversarial framework for data augmentation based on a robust optimal transport metric |
| Publication Type | Conference Paper |
| Year of Publication | 2021 |
| Authors | Su, Liyilei, Fu, Xianjun, Hu, Qingmao |
| Conference Name | 2021 IEEE 23rd Int Conf on High Performance Computing & Communications; 7th Int Conf on Data Science & Systems; 19th Int Conf on Smart City; 7th Int Conf on Dependability in Sensor, Cloud & Big Data Systems & Application (HPCC/DSS/SmartCity/DependSys) |
| Date Published | dec |
| Keywords | big data security metrics, convolutional neural network, Cosine distance, data augmentation, Deep Learning, generative adversarial network, graphics processing units, Measurement, Neural networks, object detection, Optimal Transport, pubcrawl, resilience, Resiliency, Scalability, smart cities, tensors |
| Abstract | Enhancement of the vanilla generative adversarial network (GAN) to preserve data variability in the presence of real world noise is of paramount significance in deep learning. In this study, we proposed a new distance metric of cosine distance in the framework of optimal transport (OT), and presented and validated a convolutional neural network (CNN) based GAN framework. In comparison with state-of-the-art methods based on Graphics Processing Units (GPU), the proposed framework could maintain the data diversity and quality best in terms of inception score (IS), Frechet inception distance (FID) and enhancing the classification network of bone age, and is robust to noise degradation. The proposed framework is independent of hardware and thus could also be extended to more advanced hardware such as specialized Tensor Processing Units (TPU), and could be a potential built-in component of a general deep learning networks for such applications as image classification, segmentation, registration, and object detection. |
| DOI | 10.1109/HPCC-DSS-SmartCity-DependSys53884.2021.00178 |
| Citation Key | su_convolutional_2021 |
A convolutional generative adversarial framework for data augmentation based on a robust optimal transport metric