BALTIMORE - PHOENIX - HARTFORD - BOSTON - INTERNATIONAL
The 2nd International Data Science & AI Competition
Keynote Talks
Hosted on youtube
Event Starts In
Tentatively August 30, 2020 at 9AM (New York Time)
About The Keynote Talks
The keynote talks are part of the Second International Data Science & AI Competition hosted by The International Society of Data Scientists. The keynote talks are streamed over Youtube. Participants can join from anywhere by registering below.
Location
Talks are hosted on Youtube. Register to receive the link to the conference channel
Time
Tentatively 9AM, August 30, 2020 (New York Time)
Yuchen Fama, Ph.D
Yuchen Fama is the Director of Data Science at HSB, a Munich Re company. She is responsible for developing statistical methods and large-scale machine learning solutions for engineering, insurance and IoT applications. In addition to managing a data science/engineering team, she focuses on big data, distributed computing, deep learning and cloud infrastructure. Yuchen received her Ph.D. in Statistics from the University of Connecticut.
T. Scott Clendaniel
Akshay Bahadur
Akshay is a Google Expert Developer and Software Engineer at Symantec. He is one out of 8 Google Developers Expert (Machine Learning) from India along with being one of 150 members worldwide for the Intel Software Innovator program. He is currently working alongside Google to make an Indian sign language recognition system (ISLAR) specifically aimed at running on low resource environments for developing countries.
Srivatsan Srinivasan
Srivatsan is the Chief Data Scientist at Cognizant in Phoenix, Arizona. Before Cognizant, he was the Principal Architect at Wipro Technologies. He is experienced in building complex analytical pipelines, machine learning models for extremely complex business process, petabyte scale data lake and high frequency/volume streaming analytics pipeline.
Akash Deep Singh
Akash is the COO of Tessellate Imaging. In his work, Akash is a tech wiz, passionate about solving real-world problems with artificial intelligence and machine vision. He’s worked on building novel systems to detect & classify glioma cancer and a real-time stat generation camera solution for basketball players. His past projects include autopilot firmware for search and rescue drones, building disguised and imposter face recognition software, an all-terrain navigation vehicle, and sketch to face image matching for forensics.
Nhan Tran
Mr. Nhan Tran was the Manager of Customer Data Management division at Lazada Vietnam. His mission was to help the company understand its customers behaviors by collecting hidden problems of daily transactions from data lake, interpreting it, telling its stories to the company, and bringing greater perspectives to the business. He is also a lecturer to spread his passion to the young generation of Vietnamese Data Scientist with his free time.
Abhishek Kumar Annamraju
Abhishek is the CTO at Tessellate Imaging. In his work, Abhishek is the technical lead, a computer vision consultancy and service provider. With years of expertise in a diverse set of Computer Vision challenges, working with teams from MIT Media Labs, TaTa Elxsi, Ayonix and many more. A GSoC alumnus, Abhishek has a keen interest in solving the challenges of biometric identification with publications in CVPR
Keynote Talks
Continue being updated...
Haptic Learning: Inferencing anatomical features using deep networks
Time: 9:00AM, August 30th, 2020 (New York Time)
Featured Presentation with Code
https://github.com/akshaybahadur21/ISODS-2020
Abstract: For providing haptic feedback, users have been dependent on external devices including buttons, dials, stylus, or even touch screens. The advent of machine learning along with its integration with computer vision has enabled users to efficiently provide inputs and feedback to the system. A machine learning model consists of an algorithm that draws some meaningful correlation between data without being tightly coupled to a specific set of rules. It's crucial to explain the subtle nuances of the network also the use-case we are trying to solve. The main question, however, is to discuss the need to eliminate an external haptic system and use something which feels natural and inherent to the user.
To connect the dots, we will talk about the development of applications specifically aimed to localize and recognize human features which could then, in turn, be used to provide haptic feedback to the system.
These applications will range from recognizing digits, alphabets which the user can 'draw' at runtime; developing state of the art facial recognition system; predicting hand emojis along with Google's project of 'Quick, Draw' of hand doodles. First, we will start with formulating and addressing a strong problem statement followed by a thorough literature review. Once these things are taken care of, we will discuss the data gathering part, followed by the algorithm evaluation and future scope.
Outline
The presentation will have code excerpts for the MNIST Digit Recognition and how could we use the computer vision technique to ‘draw’ digits on the screen. Using the same technique, we would also look at Quick, Draw implementation. In this, the neural network will try to recognize doodles from drawing. Subsequently, we would discuss Emojinator and the idea behind it, which will be followed by a code walkthrough and future scope for the project. Next, we would try to detect the drowsiness of the driver, which has a very strong use-case in the automobile industry. Subsequently, I would demonstrate Facial Recognition and the research paper behind the idea. Lastly, I would like to demonstrate inferencing Indian Sign Language using semantic segmentation and facial key points tracking.
While giving each of the demos, I would be talking about the model used. How to get the data and how to pre-process it. How and why to use transfer learning. Why is the literature review the most important phase of your project? How contributing to the community helps you ultimately. At the end of the session, the audience will have a clearer look at how to start building real-world projects on their own.
Demonstrations
- MNIST [10 mins]
- Quick, Draw (Google) [10 mins]
- Emojinator [10 mins]
- Drowsiness Detection [5 mins]
- Facial Recognition [5 mins]
- SignNet [5 mins]
- OpenPose [5 mins]
Getting started with Deep Learning based Computer Vision using MonkAI
Speakers: Mr. Akash Deep Singh, COO (Tessellate Imaging Private Ltd), and Mr. Abhishek Kumar Annamraju, CTO (Tessellate Imaging Private Ltd.)
Time: 9:45AM, August 30th, 2020 (New York Time)
Abstract: Deep Neural networks have shown their immense potential in their application to solving Computer Vision challenges. Highly specialized frameworks exist to build solutions across a broad spectrum of Imaging modalities from Satellite Imagery down to Microscopy. Due to the absence of workflows, the highly intricate workflows of widely used frameworks and the intrinsic complexity of the domain requires an amalgamation of mathematics and programming to build solutions.
To overcome this steep learning curve, MonkAI provides a Low code, unified wrapper across TensorFlow, Pytorch, Mxnet. Our off-the-shelf workflows allow developers to apply Deep Learning algorithms to solve Computer Vision challenges while reducing their cognitive load and entry barriers. MonkAI is an open-source toolkit with a diverse community of qualified engineers, developers and researcher across the globe pursuing a standardization of Computer Vision heuristics.
Topics of the Talk:
· Computer Vision : Past and Present (2mins)
· Deep Learning : Future (2mins)
· Skills to build while transforming from novice to expert(1min)
· MonkAI – A walkthrough(5mins)
· Hands-on exercise for applying Transfer Learning based Image Classification (15min)
· Build Deep Learning based detection, segmentation and many more Neural Networks(5mins
Deploying AI Machine Learning in Production
Speaker: Mr. Scott Clendaniel, Lead Data Scientist (Franklin Templeton), and Chief Data Officer Board of Directors, Gartner/ Evanta for DC region.
Time: 10:30AM, August 30th, 2020 (New York Time)
Abstract: 87% of all predictive models never reach production. That startling statistic is reported by both Venture Beat and Gartner, meaning that our industry has a failure rate of epic proportions. However, there are several tips and tricks for success, and this free training covers the 10 best.
Key challenges in the machine learning field, and the importance of promoting principled machine learning with more mathematical and statistical rigor
Speaker: Dr. Yuchen Fama, Director of Data Science (HSB, a Munich Re Company).
Time: 11:00AM, August 30th, 2020 (New York Time)
Abstract: In this talk we will have a brief overview of current key challenges in the machine learning field, and the importance of promoting principled machine learning with more mathematical and statistical rigor. We will also discuss fairness-aware machine learning to mitigate algorithm bias and discriminatory issues from a very high level. For more theoretical and methodological materials and discussions on certain topics, please email yuchengu@gmail.com.
Abstract: In this talk we will have a brief overview of current key challenges in the machine learning field, and the importance of promoting principled machine learning with more mathematical and statistical rigor. We will also discuss fairness-aware machine learning to mitigate algorithm bias and discriminatory issues from a very high level. For more theoretical and methodological materials and discussions on certain topics, please email yuchengu@gmail.com.
How to approach your Data Science Career in 2020?
Speaker: Mr. Srivatsan Srinivasan, Chief Data Scientist (Cognizant)
Time: 11:15AM, August 30th, 2020 (New York Time)
Abstract: Data Science is more than machine learning. Machine learning is just a small component of data science projects. There are data science projects that do not require machine learning at all as well Some job report and industry stats has resulted in plenty of college introducing machine learning and Artificial Intelligence in to their curriculum and we are flooded with online courses and data science training institutes.
End to End Machine Learning Overview
Speaker: Mr. Srivatsan Srinivasan, Chief Data Scientist (Cognizant)
Time: 11:30AM, August 30th, 2020 (New York Time)
Abstract: In this video we will be walking through an end-to-end ML Lifecyle. We will be covering ML lifecycle, starting from problem identification till model deployment and model monitoring. This video provides high level overview of different phases within ML lifecycle while I have other video in this channel detailing out individual component. For additional read on some of the phases of ML lifecycle you can read below mentioned articles.
The Second International Data Science & AI Competition - Problem, Dataset, and Rules
Time: 12:00PM, August 30th, 2020 (New York Time)
Now it is time to talk about the competition. In the this video, Mr. Nhan Tran is going to share brief information about the Second Data Science & Artificial Intelligence Competition, the data sets and the problem expected to be solved, how to participate, and some important notes from the rules.