Hi, I'm Abhishek!

I'm trying to make machines learn to think.

C++ | Python | Deep Learning | Computer Vision | Traveler
Abhishek Singh

Machine Learning Engineer

Contact Me

A little about me and what I wish to achieve:

I am a Machine Learning Engineer with an innate love for nature, a desire for traveling, and a genuine interest in meeting new people!


I am a tech-enthusiast with a firm belief in the power of technology to make a significant positive impact in the world. I love to keep myself updated with all the latest technologies and trends in the software industry.


I enjoy motivating and being part of a productive team and I am equally comfortable with working on my initiative. I have hands-on experience with Machine Learning models in a variety of problem domains using Keras and Tensorflow and other data science libraries like pandas, scikit-learn, numpy.


My research goal is to uncover the scientific principles which give rise to intelligence through learning, similar to how the human brain learns. I find Machine Learning interesting because developing our understanding of it entails developing our understanding of the principles that underlie intelligence.


Apart from this, I'm quite fond of reading books from a variety of genres.

My Projects include:

OrcaCNN: Detecting and Classifying Killer Whale from Acoustic Data

feature_project
New

Worked on 14 years of acoustic data of size 1TB in collaboration with AOOS and Axiom Data Science. Developed a deep learning model for detection of orca (or killer whale) achieving 95% accuracy on test set. Achieved 61% accuracy for classification of 20 killer whale pods. Used Google Cloud VM with GPU support for training of models.

project name

Real Time Sentiment Analysis from MemesOpen Source

A flask web-app which allows Sentiment Analysis from the web-browser from memes using the vaderSentiment tool. This web-application is mainly suitable(and made) for memes and texts that appear in social-media bodies. The motivation for this project originates from the fact that there exists no such application which could analyze the sentiment from memes. Though memes being highly contextual and sarcastic, we believe the vaderSentiment tool is of great help in this regard.

View on Github

project name

Cartpole game using OpenAI gym and DQN algorithm.Open Source

A repository sharing implemenations of Atari Games like Cartpole, Frozen Lake and OpenAI Taxi using gym. The implementations are made with DQN algortihm.

View on Github

project name

Detection of Rare Genetic Diseases using facial 2D images with Transfer LearningOpen Source

The given project leads to 98.1% accuracy and a 0.92 F1 score with results outperforming the state-of-the-art Clinical Face Phenotype Space(99.5%) for 8 classes of syndromes.

View on Github

project name

Real Time Multiple Face Recognition using Deep LearningOpen Source

A repository containing different methods for real-time multiple face recogntion using Python.

View on Github

project name

COMPUTER AIDED DIAGNOSIS OF BRAIN TUMOR USING MATLAB

The given project consists of diagnosing and detection of Brain tumour using MATLAB. (Project under the startup SketchBytes).

Find out more

project name

Medium GrabberOpen Source

This is an automated program that lets you grab the link of any article under any topic just by logging into your Google-medium account

View on Github

project name

Bootstrap Website Deployed on FirebaseOpen Source

This website shares my experience on my trekking trip to Sandakphu, the wonderland for trekkers

Website

project name

Gmail Newsletter ParserOpen Source

Get your favourite Newsletter right from your inbox to your machine. Sort out your favourites among the whole and read!

I read the newsletter from MIT Technology Review every day. What's your thing?

View on GitHub

Work Experience so far:

Associate Developer IBM (Jan 2021 - Present)

Working on Azure Cloud Microservices as an associate developer in Global Business Services (GBS).

Google Summer of Code Mentor Orcasound (March 2020 - August 2020)

Helped two students build an Active Learning pipeline (OrcaAL) which helps teach machines to detect orca (or killer whale) sounds. Specifically helped in the building, training, and pre-processing of the ML model that goes in the backend.

Link to Project

Google Summer of Code Student Developer Earth Science Information Partners (April 2019 - August 2019)

Developed two deep learning models for the detection and pod-classification of orca, or killer whale, calls in unknown long audio samples.

Link to Project

Machine Learning Research Intern Delhi Technological University (Dec 2018 - Jan 2019)

Developed a web-application for Real-Time Sentiment Analysis of Memes using flask framework in Python. The web-app was able to successfully classify memes on the basis of its polarity (Pos, Neg, Neu). The work was done under the supervision of Dr. Akshi Kumar.

Link to Project

Deep Learning Research Intern NIT Durgapur (May 2018 - June 2018)

Developed the model "Detection of Rare genetic diseases using facial 2D images with Transfer Learning" under the supervision of Dr. Dakshina Ranjan Kisku.

Link to Paper

Software Developer Intern - SketchBytes Research Labs (June 2017 - July 2017)

Developed an efficient Algortihm for Computer aided diagnosis of Brain Tumor using MATLAB.

Link to Project

Publication(s):

  • A. Singh, and D. R. Kisku, Detection of Rare Genetic Diseases using Facial 2D Images with Transfer Learning, in 2018 8th International Symposium on Embedded Computing and System Design (ISED), Kochi, India.[Link]

Extracurricular Activities:

  • Vice Chair at IEEE NIT Durgapur Student Branch for the session 2019-20.
  • Member of AquaSHIFT Team (a NASA approved project) which participated in World Challenge Finland 2018.
  • Mentor of the Project Medium_Grabber in Kaharagpur Winter of Code (KWoC) 2018.
  • Secured first position in Inter-Hall Table Tennis at NIT Durgapur - 2017
  • Secured 6th position in All India Senior School Certificate Examination-2016 among the students of Science Stream in A & N Islands.

My GitHub Acivity: