Karthik Sulthanpete Govindappa

Solving dynamic Computer Vision problems using Deep Learning, which must propel the scalability, speed and accuracy of Visual Search Services hosted by nyris GmbH.

- Research on Deep Learning in the context of Face Recognition and Object Detection - Research on Compression Technologies in the context of Convolution Neural Networks - Develop/Implement computationally efficient Deep Learning Algorithms for Embedded Systems - Develop/Optimize Machine Learning Subsystems under Flare's firmware

Master Thesis on Compressing Object Detection Convolutional Neural Networks - Network of Interest : Single-Shot MultiBox Detector - Evaluate Deep Compression Methods on aforementioned network - Results : Achieved 5x Compression rates without any accuracy loss, and up to 9x compression rates with 1-2% loss in accuracy

Enhance and optimize Real time Object Classification and Detection using CNNs

- Developed camera server and clients with GUI - Developing Python Remote Procedure Call(RPC) for distributing the computations of computer vision applications

Robot Manipulators, Mobile Robots, Computer Vision, Logic Control

Digital Communication, Signal Processing, VLSI