Deep Generative Models: How Image-Based Style Exploration Can Help Fashion Customers, Designers and Artists
Deep learning and generative models for image data have developed rapidly as research and technology in the last few years. They allow increasingly flexible creation of novel image content, and have potential to become new smart tools aiding the work of designers and artists. At the same time, images are a key aspect of fashion and e-commerce, and Zalando produces millions of fashion photos every year.
Our work leverages deep generative models and the unique image data of Zalando (paired people and product photos) in order to help both fashion designers and customers. Specifically, the talk focuses on various applications of advanced generative AI developed in our research lab. First, we are working towards novel customer-facing fashion style exploration applications. Second, we also develop controllable conditional generative models, that enable designers to iterate and interact with the neural machine, speeding-up the creative design process.
Nikolay Jetchev studied Mathematics and Computer Science at the Technical University of Darmstadt, receiving his diploma in 2007. He completed his Ph.D. in Machine Learning and Robotics at the Free University of Berlin in 2012, teaching as a postdoctoral fellow for a while. In the last several years, he has been part of the Computer Vision team at Zalando Research. Research focus: deep discriminative and generative neural networks, texture synthesis, digital image stylisation techniques, creative AI art experiments.