EyeStyle - Bring Deep Learning to Fashion Shopping
People get fashion inspiration easily from online images of various sources,e.g. social media, blogs, videos, magazines etc. EyeStyle helps online shoppers quickly find similar products from any image through a lightweight browser extension. The core technology is built around deep learning with a large product image database collected from the internet. We will introduce how we collect a large fashion database from mixed sources and how we use it to train different iterations of the algorithm that allows us to gradually improve the search accuracy, handle larger scale and increase the user experience. We will present a live demo showing how EyeStyle serves as a intelligent shopping assistant to let you purchase clothes from images in seconds.
Jie Feng is a Ph.D student at Columbia University, specializing in machine learning and computer vision. He is a member of DVMM lab and advised by Prof.Shih-Fu Chang. His research focuses on deep learning and large scale object retrieval for both 2D images and 3D models. He has interned at Microsoft, Amazon, Adobe and Google. His previous work has been granted patents and incorporated in products like Microsoft Bing and Amazon Firefly. Jie’s interest in mixing technology with other domains has lead him to found a company that build AI systems to help discover fashion products from inspiring visuals.