SAN FRANCISCO, Jan. 8, 2018 /PRNewswire/ -- Founded by two Stanford PhDs to develop deep learning solutions and enable adoption of AI at the edge, DeepVision will be demonstrating real time facial/object recognition and fine-grained object localization at CES 2018 in Las Vegas from January 9-12. The company's advanced algorithms will be featured in an FPGA-based prototype interoperating with Socionext's M10v SoC solution showcasing a new level of inference power/performance.
DeepVision brings the level of deep learning inference accuracies to the edge, same as those deep learning models in the datacenter, at much lower latency and power. Applications of the technology include security cameras, autonomous driving/ADAS, augmented reality (AR), robotics, drones and other edge applications which require ultra-low latency, high performance and low power.
Socionext, a partner of DeepVision, is a world leader in image processing. As technology advances have allowed for cloud-based functionality to become edge enabled, the need to tightly couple image sensors and convolutional neural network based hardware has become a necessity. Socionext's M10V SOC offers state-of-the-art image quality that allows for improved recognition by both humans and machines.
The DeepVision demo will be held in the Socionext Suites. Founders Dr. Rehan Hameed and Dr. Wajahat Qadeer will be on hand to answer questions and discuss AI at the edge. To sign up, please contact the Socionext Imaging Team at imaging@us.socionext.com or DeepVision Head of Business Development, Ravi Annavajjhala, at 188385@email4pr.com or 408-477-6274.
SOURCE DeepVision