Posts

Thoughts and tutorials on technology, programming, and learning:

• GlacierHack Journey My journey to winning a semantic segmentation hackathon with extreme hardware constraints.

Projects

A collection of things I've built while learning and experimenting:

• GlacierHack: Glacier Segmentation Winning solution for multi-class glacier image segmentation using UNet++ with EfficientNet-B4 backbone.
• Annotator: AI-Powered Image Annotation Web-based annotation tool with Faster R-CNN detection and Prototypical Networks for few-shot learning.
• Portfolio Website This clean, minimal website built with simple HTML, CSS, and JavaScript.

Papers

Research and academic work:

• RGFS: Reconstruction-Guided Few-Shot Learning Accepted at IGARSS 2025. A novel approach combining prototypical networks with masked autoencoding for few-shot learning on remote sensing imagery.

Info

Currently a 3rd-year Computer Science Engineering student exploring areas like Satelite-Imaging, Digital Image Processing, and Machine Learning. I believe in learning by doing and enjoy the process of turning ideas into code.

Interests: Machine Learning, Web Development, Blockchain, Digital Image Processing, Open Source Projects

Skills: Python, JavaScript, HTML/CSS, PyTorch, TensorFlow, React, Node.js, Git, Linux

Contact: Always open to discussing new opportunities, collaborations, or just chatting about technology.

Email: mohit@example.com