Follow Datanami:
March 20, 2014

Big Data Unlocks the Secret Behind Women’s Shoes

Alex Woodie

The relationship between a woman and her shoes is a deep and mysterious one. For centuries, the factors that drive a woman’s decision to acquire them have eluded researchers. Now, UK scientists claim to have cracked the biological code with a big data-powered application that tears down the wall separating a woman from finding perfect pairs of shoes for every occasion.

Cortexica yesterday announced its team of neuroscientists, visual search scientists, and machine learning engineers have come up with a novel approach to connecting women with shoes available for purchase on the Internet.

To the user, the application works simply enough. They simply snap a picture of their favorite shoes from the street, catwalk, or magazine, load it into the FindSimilar for Shoes search engine, and hit the “enter” button. Voila! They’re instantly presented with the closest matches available for purchase on the Internet.

The application appears relatively simple and easy to use on the outside, but there is much going on behind the scenes during the shoe-buying process, which scientists have long suspected. By identifying the portion of the human brain associated with a woman’s shoe-buying drive, the company has been able to replicate it with big data technology.

Cortexica’s approach uses complex “wavelet transforms” overlaid on GPUs to generate “real-time cortical key-point file descriptors” that it then mines using machine learning algorithms. This approach “mimics the spatial computation performed by biological neurons in the primary visual cortex of the human brain on a large, machine-learning scale.”

“Our team of visual search scientists, machine learning engineers, and neuroscientists reverse-engineered specific subsets of the human visual cortex and modeled the responses of neurons to create a powerful, cloud-based visual search platform that leverages visual inference engines and object recognition algorithms,” the company says on its website. “This robust visual search platform allows Cortexica to conduct highly accurate ‘exact match image search’ and ‘find similar images’ search in both still images and video.”

Fashion is a serious enterprise, and with this big data application, women can stop messing around with shoe shopping and get down to business. “Staying one step ahead of fashion is a constant challenge for many but we believe our technology helps consumers to do exactly this,” says Cortexica CEO Iain McCready. “Once the search engine is embedded in an app, the software has the ability to help shoe shoppers narrow their search and make better choices. This could result in them finding similar and often better items available elsewhere.”

Cortexica’s search technology is designed to be embedded into a mobile or Web application. In addition to shoes, it also works with handbags.

Related Items:

Lessons In Machine Learning From GE Capital

I Didn’t Know Big Data Could Do That!

MIT Spinout Exploits GPU Memory for Vast Visualization

 

 

Datanami