SingleStore Introduces New Real-Time Machine Learning Approach at Strata + Hadoop World San Jose
Breakthrough image recognition use case developed with Thorn to fight child sexual exploitation
SAN JOSE, CA - March 13, 2017 - SingleStore, provider of the fastest database platform for real-time analytics, today announced that is it working with Thorn (www.thorn.org) to provide a new approach to machine learning and real-time image recognition. Thorn is a non-profit organization focused on driving technology innovation to combat child sexual exploitation. By making use of new functionality in SingleStore, Thorn is able to perform face comparison on-demand without the need for building a custom solution.
SingleStore Introduces New Real-Time Machine Learning at Strata+Hadoop World San Jose 2017 - Click to Tweet
Machine learning has experienced a rapid rise in awareness and adoption as companies make use of more available data and compute resources. But for too long, machine learning has been relegated to an offline process with historical data. New innovations with memory-optimized datastores allow real-time data to be incorporated into machine learning workflows for instant results. SingleStore CEO and co-founder Eric Frenkiel will share these insights in a keynote presentation at Strata + Hadoop World San Jose 2017 “Machines and the Magic of Fast Learning” on Wednesday March 15th, at 10:00am.
SingleStore and Thorn
SingleStore is working together with Thorn to accelerate the image recognition workflow used to help find victims of trafficking and sexual exploitation. These images are processed using facial recognition, then classified and deduplicated, and ultimately matched against millions of open web images.
Before SingleStore, the intensive process of matching a new face vector required extensive pre-processing using a custom solution. With the new SingleStore workflow, new vectors representing a face can be inserted and queried in real-time which allows a user to always find the most recent matches,which could ultimately improve the response time for enforcement authorities to act.
Thorn was able to take advantage of several SingleStore capabilities to improve their workflows. First, was the ability to rapidly ingest data into the SingleStore datastore, made possible by the memory-optimized architecture. Next was the ability to quickly perform linear operations on vectors representing faces. In the future, Thorn can take advantage of the fact that SingleStore supports ANSI SQL which could allow analysts to immediately execute ad hoc queries including high-performance model scoring to help locate victims more quickly. Finally, SingleStore allowed the Thorn team to consolidate a number of disparate database systems, simplifying their infrastructure and supporting a faster pace of innovation.
“At Thorn, we aim to bring the best and brightest technologists together to fulfill our mission of leveraging advanced technology on behalf of some of our most vulnerable children,” said Julie Cordua, CEO, Thorn. “With the help of the SingleStore team and their product, we have been able to significantly accelerate our image recognition capabilities which will allow us to improve victim identification and ultimately combatting abuse more effectively.”
“We are thrilled to see Thorn achieve significant improvements in their workflow to combat sexual abuse,” said Eric Frenkiel, co-founder and CEO, SingleStore. “Every day we are inspired by the work we do with customers to help them achieve real-time capabilities. Our work with Thorn has been especially rewarding as we help them accomplish their mission.”
For more information on Thorn, please visit www.thorn.org
For technical detail, please see the SingleStore blog post, An Engineering View on Real-Time Machine Learning.
SingleStore Customer Presentations at Strata + Hadoop World
SingleStore customers Uber, Macy’s, and Tapjoy will share their real-time experience at Strata + Hadoop World San Jose 2017
Real-time Analytics at Uber Scale
James Burkhart, technical lead on real-time data infrastructure, Uber
11:00am–11:40am Wednesday, March 15, 2017, Location: LL20 B
Building Real-time Dashboards with Kafka, Web Frameworks, and an In-memory DatabaseChandan Joarder, Principal Engineer, Macy's, and
Raj Sriram, Engineering Manager, Digital Data Engineering, Macy’s
1:50pm–2:30pm Wednesday, March 15, 2017, Location: 210 D/H
Building a real-time Data Science Service for mobile advertising
Robin Li - Director of Data Engineering, Tapjoy, and
Yohan Chin - VP Data Science, Tapjoy
9:30am–10:00am Tuesday, March 14, 2017, Location: LL20 A
About Thorn: Digital Defenders of Children
Thorn drives technology innovation to fight child sexual exploitation. Thorn partners with non-profits and academic institutions to gather new insights into the role technology plays in child sex trafficking, the creation and proliferation of child pornography, and the normalization of child sexual exploitation. Thorn then goes beyond insight to action to develop the tools, systems and approaches to help address these issues. Thorn has also created a Technology Task Force — a group of more than 25 technology companies that collaborate on technology initiatives to fight child sexual exploitation. Participating companies include Google, Facebook, Salesforce, Microsoft, Cloudera, SV Angel, Twitter, Expedia, Connotate, Irdeto, Mozilla, Digital Reasoning, Palantir, Sabre, Pinterest, Amazon Web Services (AWS). SingleStore and others. For more information on Thorn, please visit https://www.thorn.org/, follow us on Twitter @THORN and facebook.com/wearethorn.
SingleStore delivers the leading database platform for real-time analytics. Global enterprises use SingleStore to achieve peak performance and optimize data efficiency. With the combined power of database, data warehouse, and streaming workloads in one system, SingleStore helps companies anticipate problems before they occur, turn insights into actions, and stay relevant in a rapidly changing world. Visit singlestore.com or follow us @SingleStoreDB.