Presenter: Farhad Fallahlalehzari, Applications Engineer, Aldec
Thursday, October 31, 2019
Machine Learning (ML) is becoming a fundamental part of almost any computer vision-based application on the edge. From pedestrian detection in ADAS to cancer diagnosis in medical, and quality assurance in agriculture. However, there are challenges involved in developing an optimized and high-precision machine learning applications on the edge such as selecting the right processing system and neural network. FPGAs, as edge computing units, have shown a solid potential on improving the performance of ML applications. Aldec has recently developed DNN-based object detection applications on TySOM-3A-ZU19EG embedded development board (using Xilinx® Zynq™ MPSoC™ FPGA) for its customers to kick start their ML projects. In this webinar, you will learn about the ML application development process and what tools are required to simplify the design and implementation for FPGA-based machine learning applications.
- AI vs machine learning vs deep learning
- Deep learning application development flow and its challenges
- ASICs vs GPUs vs FPGAs - which one is better for deep learning applications?
- Main methods to achieve high-performance machine learning applications
- How to develop an DNN-based object detection application using Xilinx Zynq MPSoC FPGA on TySOM embedded development board