Presented By O’Reilly and Intel AI
Put AI to Work
April 29-30, 2018: Training
April 30-May 2, 2018: Tutorials & Conference
New York, NY

The quest for a new visual search beyond language

Mike Ranzinger (Shutterstock)
4:00pm–4:40pm Wednesday, May 2, 2018
Models and Methods
Location: Grand Ballroom West

Who is this presentation for?

  • Production engineers, AI practitioners, and researchers interested in the future of visual search

Prerequisite knowledge

  • A general understanding of convolutional neural network architectures, including the role of individual components of the networks, and deep learning NLP techniques, such as word embeddings and word2vec

What you'll learn

  • Explore the research behind Shutterstock's composition-aware search AI technology


Words are no longer sufficient in delivering the search results users are looking for, particularly in relation to image search. Text and languages pose many challenges in describing visual details and providing the necessary context for optimal results. Machine learning technology opens a new world of search innovation that has yet to be applied by businesses.

Mike Ranzinger shares his research on composition-aware search and explains how the research led to the launch of AI technology that allows Shutterstock’s users to more precisely find the image they need within the company’s collection of more than 150 million images. This technology makes heavy use of modern machine vision, natural language processing, and information retrieval techniques. While the company released a number of AI search-enabled tools in 2016, this research and technology expands on existing tools to identify the networks that localize and describe regions of an image as well as the relationships between things.

Photo of Mike Ranzinger

Mike Ranzinger


Mike Ranzinger is a senior research engineer at Shutterstock, where he and a team of researchers and engineers have invented a number of AI search technologies and collaborated on multiple patent filings. Previously, Mike held a variety of software developer roles at New Century Software, Boulder Imaging, and AlchemyAPI (acquired by IBM Watson), where he spearheaded a natural scene optical character recognition (OCR) project that provided an API to extract text from images and was a member of the larger machine vision group that launched the industry’s first commercial image tagging and similarity API. Mike first became enamored with ray tracers and machine vision while studying at Colorado State University. Mike is passionate about cycling and spends most of his free time training for races as a new domestic pro. He holds a BS in computer science from Colorado State University.