Description:

In this workshop, we will try to teach how to understand Deep Learning, various paths to follow, Domains to explore and the most important part- how to start with the paper selection and implementation. We will also learn how to deploy a simple model into production.

This workshop aims at providing the attendees of all level a foundation of research and further prospectives in deep learning.

Contents

• Paths and prospects in Industry and Academia (10 minutes)
• Difference between AI, ML, and DL. (5 minutes)
• Introduction to Deep Learning frameworks (Hands-on) (5 minutes)
• Paper selection (10 minutes)
• Implementation (Hands-on) (60 minutes)
  • Understanding the dataset
  • Feature Extraction
  • Model Selection
  • Data Formatting
  • Comparison
• Demonstration of our work (General Overview)
  • Audio Tagging
  • Acoustic scene classification
  • Visual Question Answering
• Publish/Deploy (Hands-on) (30 minutes)
• Stay Motivated
• Opportunites to explore

The participants should have interest in Research. Detailed examples of practical applications and the scope of deep learning are found throughout the presentation. However, this presentation may prove difficult for first-time learner’s of AI to comprehend.

Prerequisites:

Preferred

• Basic Python concepts
• Basic knowledge about Machine Learning Algorithms.

Preferred (but not necessary)

• Interest in working on Research problems

Installed libraries:

• Keras
• Theano or Tensorflow

Content URLs:

Content URLs:

Github reposistories:

• Keras_aud
• Audio-Vision

Drive links:

• Content link : (Slides to be uploaded soon)

Speaker Info:

Aditya Arora and Akshita Gupta are currently final year semester exchange students at Indian Institute of Technology, Roorkee. They have been working on research problems using deep learning specifically in Audio processing and visual Q&A. Aditya is a member of various open source societies such as rust-community while Akshita has experience in Academia research and is a selected as an Outreachy intern at Mozilla 2018. They have been working in python for the past 4 years and have been moving forward working on Computer Vision and Audio processing problems.

Speaker Links:

• Twitter : https://twitter.com/imaarora
• Twitter : https://twitter.com/akshitac8
• Linkeldn: https://linkedin.com/in/aditya-arora145/
• Linkeldn: https://www.linkedin.com/in/akshita-gupta152/
• Github : https://github.com/channelcs
• Blog : http://channelcs.github.io/