About⌗
Hi there! I’m Arindam Ghosh, a mathematician turned computer vision researcher and co-founder @Acquora. My passion lies in applying AI to solve deep tech hard problems in healthcare and space tech.
My team made it to the top 0.5% in NASA Space Apps 2022, and our AI solution won first prize at the Smart India Hackathon (ISRO even uses it now!). I also got to inspire young minds by organizing a science hackathon for high schoolers in Barcelona, Spain.
At Acquora, we're using geospatial data for sustainable water management in lithium mining and mine tailings. It's all about using space tech to solve Earth's problems.
When I'm not working, I'm either geeking out over new tech, struggling through piano lessons 🎹, or picking up my paintbrush after a 10-year break. Wish me luck!
Skills⌗
- Image Processing and Computer Vision using OpenCV, TensorFlow, PyTorch
- Medical Image Analysis and for biomedical image applications
- Training and deploying Deep Learning models
- Geospatial Data Analysis with QGIS, GDAL, GeoPandas
- Created large dataset of geospatial satellite imagery using INSAT-3D/3DR data
- Cloud Computing with AWS, experience with PySpark, SQL
- Web Development with React
- Python Programming for computer vision, data analysis, machine learning
- C/C++ Programming for performance-critical vision applications
- Parallel Programming with CUDA for GPU optimization
View all my projects and experiences by tag to find something more specific.
[Research Projects]
$ Neural Spike Detection in Retina Microscopy
Exploratory analysis of spike detection techniques for light sheet microscopy images of organoids in photoreceptor restoration therapy research.
$ Optimizing Proton Tracking and Energy Determination for Improved Proton CT Imaging
Developing algorithms using image processing and ML to reconstruct proton energy by analyzing trajectories through range detectors.
$ Brain Tumor MRI Classification using Deep Learning
Using LBP, HOG feature descriptors and transfer learning models (VGG-16, ResNet-18, DenseNet-121) for MRI brain tumor classification.
$ Review of Brain Tumor Segmentation Methods
Comparative study of ML and CV methods for brain tumor segmentation using BraTS challenge dataset (2012-2020).
$ INCYDE: Large Scale Cyclone Detection and Intensity Estimation Dataset
Large-scale cyclone detection and intensity estimation dataset using satellite infrared imagery (Currently under review).
$ Cyclone Vision: Deep Learning Framework and Web App
Comprehensive deep learning framework and web application for early cyclone detection and monitoring (Currently under review).
$ HEp-2 Cells Classification using Dictionary Learning
Comparative study of discriminating Dictionary Learning methods for classifying HEp-2 cells staining patterns.
$ Generative Adversarial Networks (GAN) to Denoise Images
Training GAN models to denoise images using ImageNet ILSVRC2013 dataset with added Gaussian noise.
$ Forex Forecasting Using Support Vector Machines
Mathematical implementation of SVC, SVR and Kernels for Forex currency pair forecasting with kernel comparison.
$ Parallel Programming using CUDA
Boolean matrix-vector multiplication (BMV) using Wiedemann method implemented in CUDA at HPCL.
[Weekend Projects]
[Hackathons & Competitions]
$ NASA SpaceApps Challenge 2022
Top 0.5% finalist among 31,561 participants worldwide. Developed innovative solution for measuring open science.
$ Smart India Hackathon 2022
Won national hackathon solving ISRO's problem statement on cyclone intensity estimation using deep learning.
$ #T-TeC 2023 - Telespazio Technology Contest
Won 1st place in Italy's premier space technology contest with ESTESIA project, awarded €10,000 and accelerator invitation.
$ Google CodeJam 2020
Successfully qualified through the Google CodeJam Qualification Round.
$ Google Hash Code 2020
Received Google Hash Code certification for successfully completing the qualification round.
[Blog Posts]
$ Start here for a quick overview of everything you need to know
We've crammed the most important information to help you get started with Ghost into this one post. It's your cheat-sheet to get started, and your shortcut to advanced features.
$ Customizing your brand and design settings
How to tweak a few settings in Ghost to transform your site from a generic template to a custom brand with style and personality.
$ Writing and managing content in Ghost, an advanced guide
A full overview of all the features built into the Ghost editor, including powerful workflow automations to speed up your creative process.
$ Building your audience with subscriber signups
How Ghost allows you to turn anonymous readers into an audience of active subscribers, so you know what's working and what isn't.
$ Selling premium memberships with recurring revenue
For creators and aspiring entrepreneurs looking to generate a sustainable recurring revenue stream from their creative work, Ghost has built-in payments allowing you to create a subscription commerce business. Connect your Stripe account to Ghost, and you'll be able to quickly and easily create monthly and yearly premium plans for members to subscribe to, as well as complimentary plans for friends and family. Ghost takes 0% payment fees, so everything you make is yours to keep! Using subscrip
$ How to grow your business around an audience
A guide to collaborating with other staff users to publish, and some resources to help you with the next steps of growing your business
Contact⌗
Feel free to reach out to me via these platforms: