Try Interactive Portfolio

Akanksha Mhadolkar

Computer Vision and Deep Learning Researcher

About Me

I'm a Computer Vision and Deep Learning Researcher with a passion for creating innovative solutions. Currently working as a Junior Research Fellow at IIT Hyderabad, I specialize in computer vision, deep learning, and 3D reconstruction techniques. My research focuses on applying advanced computer vision methods to heritage structure analysis and reconstruction.

Skills

Programming Languages

Python, R, SQL, HTML/CSS

ML Frameworks and Libraries

TensorFlow, PyTorch, Scikit-learn, Keras, OpenCV, Yolov5, HuggingFace, Transformers, Langchain, Llama

Development Tools

VS Code, IntelliJ IDEA, PyCharm, Blender, QGIS, Jupyter Notebook

Web Development

React, Node.js, Three.js

Work Experience

Junior Research Fellow

Indian Institute of Technology Hyderabad

November 2022 - Present

  • Applied advanced computer vision methods for point cloud semantic segmentation, unraveling the architectural intricacies of heritage structures.
  • Pioneered 3D reconstruction of heritage sites using image-based techniques like Visual SFM, COLMAP, and NeRF

Summer Research Intern

Indian Institute of Management Indore

June 2022 - August 2022

  • Worked primarily on road accidents data in India and USA and performed times series forecasting using ARIMA/SARIMA to predict future road accidental deaths.

Artificial Intelligence Research Intern

Utkarshini Edutech

January 2022 - April 2022

  • Used Explainable AI to explain predictions of an image classification model.
  • Explored different causal machine learning libraries and implemented the same to understand the cause and effect of telecom customer churn

Projects

Thing Finder

A web application that allows users to log, track, and manage the whereabouts of personal items.

EyeSpeak

A python application that uses deep learning models to track eye movements to type using a virtual keyboard for patients with various motor neuron diseases.

Image Colorization

Colorize grayscale images with autoencoders.