One of the biggest problems when shopping online for apparel and footwear is choosing the correct size and determining a good ‘fit’ without try-on. This fails many times and the consumer is left frustrated and disappointed, while the retailer is left with the cost and logistical headache of returns. The talk will discuss how modern Computer Vision and Data Science techniques can be used to address this problem at scale, with emphasis on footwear. Different technological choices can vary widely in how accurate and robust they are vs. their ability to scale and ease-of-use. Finding a sweet spot in this trade-off is tricky. Nike has recently begun deploying computer vision and deep learning based products to solve this problem. We’ll dive deeper into one of these solutions to discuss technical challenges in its development and some of its pros and cons.
Want to sponsor? Contact us to find out more.