Keynote: Towards Large-scale Deployment of Intelligent Video Analytics Systems

Dr. Julien Lai | Director of the Applied DL team of NVIDIA APAC Region

Abstract: One of the grand challenges of AI is to understand video content. Applications are endless: video surveillance, self-driving cars, live video streaming, ad-placement, etc. Problem is that Deep Learning, which has been boosting modern AI, is computationally expensive. It’s even more challenging when it comes to live stream video. Chip manufacturers are continuously pushing the edge of current IC technology to achieve both high throughput/performance, and high power efficiency. Engineers have also applied techniques such as network pruning, low precision, and model compression, to further accelerate network inference. That is also why our team are building the NVIDIA DeepStream SDK, which simplifies development of high performance video analytics applications powered by deep learning. We have seen successful large-scale deployment of such intelligent video analytics systems, and we see this as an unstoppable trend.

jlaiBio: Dr. Junjie Lai received his bachelor and master degrees from Tsinghua University, and received his PhD degree from INRIA, France. His PhD research focused on GPU architecture study, performance analysis and optimization.

Dr. Lai is currently the director of the HPC software development and Applied Deep Learning team of NVIDIA APAC region. Besides leading the team, he is collaborating with developers from many well-known internet companies, to better accelerate their deep learning applications with NVIDIA GPUs. His primary focus now is to apply Deep Learning/Machine Learning techniques for various real-world problems in Computer Vision, Finance, etc.