Perl @ BASH, Level 3
79 Ayer Rajah Crescent
Early Bird [Until 10 May 19] (Ticket Inclusive of G.S.T) - $457.43
Normal Ticket (Ticket Inclusive of G.S.T) - $481.50
Fundamentals of Deep Learning for Multiple Data Types
Presented by SGInnovate, NVIDIA & NSCC
Learn to build Deep Learning applications for image segmentation, sentence generation, as well as image and video captioning. You should also have successfully completed the ‘Fundamentals of Deep Learning for Computer Vision’ DLI course or an equivalent.
Led by experienced instructors, you can expect to learn the latest deep learning techniques for designing and deploying neural network-powered machine learning across a variety of application domains.
In the course, participants will learn to:
- Implement common deep learning workflows such as image segmentation and text generation
- Compare and contrast data types, workflows, and frameworks
- Combine deep learning powered computer vision and natural language processing to start
- Solve sophisticated real-world problems that require multiple input data types
This workshop teaches you to apply deep learning techniques to a range of problems involving multiple data types through a series of hands-on exercises. You will work with widely-used deep learning tools, frameworks, and workflows by performing neural network training on a fully-configured GPU accelerated workstation in the cloud.
After a quick introduction to deep learning, you will advance to building deep learning applications for image segmentation, sentence generation, and image and video
captioning, while simultaneously learning relevant computer vision, neural network, and natural language processing concepts. You should also have successfully completed the ‘Fundamentals of Deep Learning for Computer Vision’ DLI course or an equivalent.
At the end of the workshop, you will be able to assess a broad spectrum of problems where you can apply deep learning.
Click here for more SGInnovate – NVIDIA Training Programmes.
08:45am – 09:00am: Registration
09:00am – 09:45am: Introduction
Introduction to deep learning, situations in which it is useful, key terminology, industry trends and challenges
- Course Overview
- Getting Started with Deep Learning
09:45am – 10:00am: Tea Break
10:00am – 12:00pm: Image Segmentation with TensorFlow
Hands-on exercise: Segment MRI images to measure parts of the heart using tools such as TensorBoard and the TensorFlow Python API
- Compare image segmentation to other computer vision problems
- Experiment with TensorFlow tools
- Implement effective metrics for assessing model performance
12:00pm – 1:00pm: Lunch
1:00pm – 3:00pm: Word Generation with TensorFlow
Hands-on exercise: Train a Recurrent Neural Network to understand both images and text, and to predict the next word of a sentence using the MSCOCO (Microsoft Common Objects in Context) dataset
- Introduction to Natural Language Processing (NLP) and Recurrent Neural Networks (RNNs)
- Create network inputs from text data
- Test with new data
- Iterate to improve performance
3:00pm – 3:15pm: Tea Break
3:15pm – 5:15pm: Image and Video Captioning
Hands-on exercise: Train a model that generates a description of an image from raw pixel data by combining outputs of multiple networks (CNNs and RNNs) through concatenation and/or averaging.
- Combine computer vision and natural language processing to describe scenes
- Learn to harness the functionality of Convolutional Neural Networks (CNNs) and RNNs
5:15pm – 5:30pm: Summary
Review of concepts and practical takeaways
- Summary of key learnings
- Workshop survey
- Regular Ticket: SGD$450 / pax (before GST)
- 5% Early Bird Discount (ends 10 May 2019) - SGD$427.50 / pax (before GST)
For group discounts or enquiries, please contact [email protected]
Huaizheng Zhang is a third-year PhD student at the School of Computer Science and Engineering, Nanyang Technological University (NTU).
His research interests are in Deep Learning, Machine Learning, and video analysis. He has published papers at the International Conference of Data Mining (ICDM), the International Conference of Communications (ICC), and on Transfer Matrix Method (TMM) and Integrated Computational Materials Engineering (ICME).
He has also given presentations as a subject matter expert. He is an expert in PyTorch and Tensorflow and enjoys sharing his Deep Learning code on GitHub. Besides the field of research, he is also an experienced educator and helps students at NTU apply deep learning in many courses as a teaching assistant.