Anurag Paul

Anurag Paul

San Francisco Bay Area
966 followers 500+ connections

About

Experienced Machine Learning Engineer with a demonstrated history of working in the…

Activity

Join now to see all activity

Experience

  • Plus Graphic

    Plus

    Santa Clara, California, United States

  • -

  • -

    Greater San Diego Area

  • -

    Cupertino, California

  • -

    Greater San Diego Area

  • -

    Gurgaon, India

  • -

    Gurgaon, India

  • -

    New Delhi Area, India

  • -

    Bangalore

  • -

    Duisburg Area, Germany

Education

  • UC San Diego Graphic
  • -

    Activities and Societies: Information Management Group (IMG), Literary Section (Debating)

    Received Silver Medal from IIT Roorkee for scoring highest CGPA in the Production and Industrial Engineering Batch of 2012.

    Leadership Roles:
    • Executive Member, Students Affairs Council, IIT Roorkee
    • Coordinator, Information Management Group, IIT Roorkee –
    • Manager(Web), Cognizance 2011 – the technical festival of IIT Roorkee
    • Joint Secretary, Literary and Debating Society, IIT Roorkee

Licenses & Certifications

Publications

  • Efficient Transformer Attention for GenAI

    DZone AI/ML

    Explore this overview of recent research focused on building efficient transformer models by reducing the computational complexity of self-attention.

    See publication
  • Unboxing the Black Box

    DZone AI/ML

    Explore explainable AI by unpacking heat maps and Structured Attention Graphs for explaining model decisions in computer vision applications.

    See publication
  • Practitioner’s Guide to Deep Learning

    DZone AI/ML

    An overview of best practices, recommendations, and fundamentals for managing a deep learning project to ensure a robust AI product.

    See publication
  • Characterizing Joint Attention Behavior during Real World Interactions using Automated Object and Gaze Detection

    ACM ETRA 2019

    Joint attention is an essential part of the development process of children, and impairments in joint attention are considered as one of the first symptoms of autism. In this paper, we develop a novel technique to characterize joint attention in real time, by studying the interaction of two human subjects with each other and with multiple objects present in the room. This is done by capturing the subjects' gaze through eye-tracking glasses and detecting their looks on predefined indicator…

    Joint attention is an essential part of the development process of children, and impairments in joint attention are considered as one of the first symptoms of autism. In this paper, we develop a novel technique to characterize joint attention in real time, by studying the interaction of two human subjects with each other and with multiple objects present in the room. This is done by capturing the subjects' gaze through eye-tracking glasses and detecting their looks on predefined indicator objects. A deep learning network is trained and deployed to detect the objects in the field of vision of the subject by processing the video feed of the world view camera mounted on the eye-tracking glasses. The looking patterns of the subjects are determined and a real-time audio response is provided when a joint attention is detected, i.e., when their looks coincide. Our findings suggest a trade-off between the accuracy measure (Look Positive Predictive Value) and the latency of joint look detection for various system parameters. For more accurate joint look detection, the system has higher latency, and for faster detection, the detection accuracy goes down.

    See publication

Patents

Courses

  • Advanced Statistical NLP

    CSE 291 E

  • Deep Learning for Sequences

    CSE 291G

  • Linear Algebra

    ECE 269

  • Machine Learning for Image Processing

    ECE 285

  • Probability and Statistics for Data Science in Python

    ECE 225A

  • Statistical Learning 1

    ECE 271A

  • Statistical learning 2

    ECE 271B

Projects

  • Domain Adaptation using Generative Adversarial Networks

    -

    Developed a system to generate realistic images from synthetic dataset to create more training data for self-driving cars. For this, we developed and compared the performance of Cycle GANs and Dual GANs in adapting images from GTA V to Cityscapes dataset.

    Other creators
    See project

Honors & Awards

  • Kaizen

    Delhivery

    For exceptional work in developing a facility location planner.

  • Star Novitiate

    iRunway India Pvt. Ltd.

    For outstanding performance and lasting contribution

  • Silver Medal

    Indian Institute of Technology Roorkee

    For scoring the highest CGPA among the graduating students of my batch

  • DAAD Scholarship

    German Academic Exchange Service

    Research Scholarship for summer internship in Germany

Recommendations received

2 people have recommended Anurag

Join now to view

More activity by Anurag

View Anurag’s full profile

  • See who you know in common
  • Get introduced
  • Contact Anurag directly
Join to view full profile

Other similar profiles

Explore collaborative articles

We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.

Explore More

Others named Anurag Paul in United States

Add new skills with these courses