ICADS '22 Talk: Fuzzy Loss Functions for Generative Adversarial Networks

26 views
Download
  • Share
Create Account or Sign In to post comments
#Generative Adversarial Network (GAN) #Loss Function #Fuzzy Logic #Medical Image Analysis #Covid-19 #Deep Learning

Generative Adversarial Networks (GAN) are very popular for medical image analysis. This talk presents an innovative fuzzy loss function for the GAN in the domain of image analysis.

The GAN architecture presented here uses two convolutional neural networks, one of which is a generator and the other is a discriminator. Besides the loss function, the training algorithm is also presented in the paper. The proposed approach is generic and can be used for many domains pertaining to medical image analysis and diagnosing.

The proposed work with the modified loss function is experimented on the Covid-19 image set. The results achieved are discussed in brief. In the end, the paper presents limitations and future enhancements possible based on the work.

Speaker Professor Priti S. Sajja delivered this presentation during the IEEE Computer Society, Santa Clara Valley Chapter’s International Conference on Applied Data Science (ICADS'22) on July 12, 2022, virtually.

For access to past video webinars or to join our Dlist to hear about future programs, please visit https://r6.ieee.org/scv-cs

Generative Adversarial Networks (GAN) are very popular for medical image analysis. This talk presents an innovative fuzzy loss function for the GAN in the domain of image analysis.

The GAN architecture presented here uses two convolutional neural networks, one of which is...

Speakers in this video

Advertisment

Advertisment