Create your own Devfolio?
DhruvAgarwal

Dhruv Agarwal

I am a machine learning enthusiast pursuing bachelors in Information Technology from IIIT-Allahabad

Projects

See++

A Dynamic Real-Time visual aid for blind peopleAndroid, OpenCV, Internet of Things (IoT), Deep Learning, Natural language processing (NLP), Computer Vision

ShareHub

We bring you a secure platform where you can find similar people, with whom you can split the fee and not worry about fraud or that account being shared further with anyone else.React, JavaScript, Flask, React-redux

Skills

Artificial Intelligence
Android
Machine Learning
Computer Vision
Natural Language Processing

Experience

  • IIT Delhi - Summer Research Fellow
    May 2019 - July 2019

    Summer Research Fellowship Programme (SRFP) 2019, Indian Institute of Technology Delhi, Delhi

    Worked on the project Mobility Assistant for Visually Impaired (MAVI), which is aimed at enabling mobility for visually impaired individuals.

    Developed a Deep Learning based object detection model to detect cows in an image.

    Carried out research on estimating uncertainty in state-of-the-art object detection algorithms.

    Formulated deep learning based approaches to score uncertainty estimate of object detection algorithms.

  • SAP Labs - SDE Intern
    June 2020 - August 2020

    Worked with Digital Manufacturing Cloud team to build different skills for SAP Conversational AI (CAI) chatbots:

    • Trained SAP CAI FAQ bot on a dataset of manually prepared question-answer pairs from the DMC FAQ page.
    • Developed a new skill for SAP CAI bot to generate analytical insights into a team’s/ individual’s contribution. A pipeline of python programs was developed and deployed to SAP Cloud Foundry and made available to the bot through an API endpoint.

  • RWTH Aachen University - Research Intern
    June 2020 - August 2020

    Worked closely with a research team on the project, “MessAgE”, where we developed machine learning techniques to
    migrate measurement uncertainties from one measurement process to another.

    • Proposed a machine learning-based method, combining inclusive similarity and family similarity of
    measurement processes to transfer measurement uncertainties. The proposed algorithm achieved the best results when compared with already existing algorithms.

  • Bielefeld University - Research Intern
    August 2020 - Present

    Working on, PHYRE, an ongoing team project with Dr. Andrew Melnik, University of Bielefeld, Germany, where
    we are developing learning-based approaches to simulate physics by exploring interactions amongst objects and the environment using computer vision and deep learning.