Chulalongkorn Uni: Intro to Deep Learning with NVIDIA GPU in Computer Vision
A COLLABORATION WITH CHULALONGKORN UNIVERSITY Computer Vision - Workshop Dates: DEC 18 - 20 (3 days) *OPEN FOR REGISTRATION Language of Conduct:Thai language & English Materials Course Fee:Regular Price : USD900 Early bird for corporate : USD600 (by July 1st , 2019)Early bird for academic : USD550 (by July 1st , 2019)Price EXCLUDES 7% VAT. Full price will be included in the ticket. *Academic tickets MUST be purchased and proven with a EDU/University email! HOW TO REGISTER: 1. To register directly with iTrain Asia, please email your details to info@itrainasia.com2. Via Eventbrite registration page*Please note that Eventbrite charges apply PAYMENT METHODS: 1. Via Eventbrite by PayPal/Credit Card (please note that Eventbrite charges will apply) 2. To make direct payment transfer, please transfer to the following account and kindly SEND YOUR FULL NAME & RECEIPT upon successful payment: Please key in the price Please email your receipt/proof of payment to yana@itrainasia.com Pre-requisites Must have technical knowledge in R and Python, understand basic Data Science, Machine Learning and AI algorithms, familiarity with basic programming fundamentals such as functions and variables Your Certificate You will receive an e-Certificate by NVIDIA Deep Learning Institute upon completion. About the Course This workshop teaches you to apply deep learning techniques to a range of computer vision tasks through a series of hands-on exercises. You will work with widely-used deep learning tools, frameworks, and workflows to train and deploy neural network models on a fully-configured, GPU accelerated workstation in the cloud. After a quick introduction to deep learning, you will advance to building and deploying deep learning applications for image classification and object detection, followed by modifying your neural networks to improve their accuracy and performance, and finish by implementing the workflow that you have learned on a final project. At the end of the workshop, you will have access to additional resources to create new deep learning applications on your own. Learn the latest techniques on how to design, train, and deploy neural network-powered machine learning in your applications. You’ll explore widely used open-source frameworks and NVIDIA’s latest GPU-accelerated deep learning platforms. DLI Workshop Attendee Instructions: You MUST bring your own laptop to this workshop. WHAT YOU WILL LEARN Course Outline Introduction to Deep Learning with NVIDIA GPU in Computer Vision 3 days from 9AM - 4:30PM DAY 1 - (DEC 18, 2019) Platform: Keras on Google Colab Prerequisite: Python programming What is Deep Learning and what are Neural Networks? (90 mins) [Lecture] Basics of Deep Learning Training a Neural Network Practical session I (90 mins) [Lab] Create a Neural Network in Python Introduction to convolution neural networks and recurrent neural networks (90 mins) [Lecture] Intuition and building blocks Types of convolutional neural networks Types of recurrent neural networks Practical session II (60 mins) [Lab] Convolutional Neural Networks and Recurrent Neural Networks Tips and tricks to training a neural network model (30 mins) [Lecture] DAY 2 [DLI] (DEC 19, 2019) Platform: DIGITS NVIDIA Deep Learning Institute Fundamentals Training Pre-requisite: MUST have technical background and basic understanding of Deep Learning concepts Certificate: Participants will receive an e-certificate from Deep Learning Institute Image Classification with DIGITS (120 min) How to leverage deep neural networks (DNN) within the deep learning workflow Process of data preparation, model definition, model training and troubleshooting, validation testing and strategies for improving model performance using GPUs. Train a DNN on your own image classification application Object Detection with DIGITS (120 min) Train and evaluate an image segmentation network Neutral Network Deployment with DIGITS and TensorRT (120 min) Uses a trained DNN to make predictions from new data Show different approaches to deploying a trained DNN for inference learn about the role of batch size in inference performance as well as virus optimisations that can be made in the inference process DAY 3 (DEC 20, 2019) Intelligent Video Analytics with Deep Learning Platform: Keras on Google Colab Prerequisite: Python programming & 1st training day Overview of Architectures for Computer Vision (90 min) Lab 1: Image classification with Keras (60 min) Deployment with Deepstream and TensorRT (30 min) Lab 2: Deployment for classification and detection tasks (30 min) Transfer learning techniques (30 min) Lab 3-1: Model adaptation (30 min) Lab 3-2: Advanced techniques for adaptation (30 min) Video action recognition (15 min) Lab 4: Video action recognition (30 min) Jetson Demo: deployment on Jetson (15 min) ABOUT YOUR TRAINERS: PROF. EKAPOL CHUANGSUWANICH, Ph.D. Ekapol Chuangsuwanich is a Faculty Member in the Department of Computer Engineering at Chulalongkorn University. He received the B.S. and S.M. degree in Electrical and Computer Engineering from Carnegie Mellon University in 2008 and 2009, respectively. He then joined the Spoken Language Systems Group at MIT Computer Science and Artificial Intelligence Laboratory. He received his Ph.D. degree in 2016 from MIT. His thesis work was on low-resource automatic speech recognition and representation learning using neural networks, which was part of the system that won Babel open keyword spotting challenge in 2016. With his expertise in multimedia retrieval, he is also one of the founding members of SmartVid.io, a startup working on organizing videos and images for the construction industry. In 2017, SmartVid.io was a runner-up in NVIDIA's Inception competition for AI startups. PROF. PEERAPON VATEEKUL, Ph.D. Peerapon Vateekul received his Ph.D. degree from Department of Electrical and Computer Engineering, University of Miami (UM), Coral Gables, FL, U.S.A. in 2012. Currently, he is an assistant professor at Department of Computer Engineering, Faculty of Engineering, Chulalongkorn University, Thailand. Also, he is a deputy head of the department in academic affairs. His research falls in the domain of machine learning, data mining, deep learning, text mining, and big data analytics. To be more specific, his works include variants of classification (hierarchical multi-label classification), data quality management, and applied deep learning techniques in various domains, such as, medicinal images and videos, satellite images, meteorological data, and text.
read more