Yash Bhalgat

I am PhD student with the Visual Geometry Group (VGG) at the University of Oxford. Previously, I was a Research Scientist at Qualcomm AI Research, where I did algorithm and system design to develop efficient deep networks for computer vision usecases. I have developed pipelines and published papers on low-bit quantization, structured and unstructured pruning of deep networks. The usecases I have worked on include object detection and classification, video segmentation, gaze estimation, etc. My current research focuses on developing compute-adaptive deep networks. I previously completed my Masters from University of Michigan, Ann Arbor and Bachelors from IIT Bombay. I have worked at IBM Research - Almaden, IBM India Research Lab, Tata Research Center and Infurnia (a startup company based in Mumbai). SKILLS: Python, C++, C, SQL, Julia, MATLAB, R, PyTorch, TensorFlow, Keras, OpenAI gym, Theano, CUDA, git For more information, please visit my webpage: https://yashbhalgat.github.io