Get in Touch

Thank you for your interest! Please fill out the form below if you would like to work together.

Thank you! Your submission has been received!

Oops! Something went wrong while submitting the form

My Portfolio

Tensorflow Object Detection

Introducing Tensorflow-Object-Detection - a powerful computer vision project that uses Python and Tensorflow to detect and recognize objects in real-time video streams.

This project is built on top of a pre-trained model that has been fine-tuned to detect objects with high accuracy. By modifying the model to run on a video stream, this project can recognize and track multiple objects simultaneously, providing users with real-time object detection and tracking capabilities.

Tensorflow-Object-Detection is a perfect tool for applications such as security and surveillance, autonomous vehicles, and robotics, where accurate object detection and tracking are critical to the system's performance.

If you're looking to dive into the world of computer vision or want to add real-time object detection and tracking to your project, then Tensorflow-Object-Detection is an excellent place to start. With its easy-to-use Python interface, powerful Tensorflow backed, and real-time video streaming capabilities, this project has everything you need to get started with object detection and tracking.

So, if you're interested in exploring the exciting world of computer vision, head over to the Tensorflow-Object-Detection project and start detecting objects today !

Stable Diffusion - Image to Prompts

Currently working on the Stable Diffusion - Image to Prompts project.The goal of this project is to build a model that can predict the text prompt that generated a given image, instead of the traditional approach of generating an image from a text prompt.

By building a model that can predict the prompt from an image, you will gain insights into the latent relationships between prompts and images, and learn how to reliably invert the diffusion process that generated the image.

Real Time Sign Language Detection

Introducing Real-time Sign Language Detection - a powerful machine learning project that uses Python and Tensorflow to detect and recognize sign language gestures in real-time video sequences.

This project uses the MediaPipe Holistic Keypoints API to extract key body points from video frames, and builds a Sign Language model using Action Detection powered by LSTM layers. With this powerful machine learning model, the project can accurately predict sign language gestures in real-time video sequences.

Real-time Sign Language Detection is a perfect tool for applications such as communication with hearing-impaired individuals, language learning, and research in the field of sign language. With its easy-to-use Python interface, powerful Tensorflow backed, and real-time video streaming capabilities, this project has everything you need to get started with sign language detection and recognition.
‍
So, head over to the Real-time Sign Language Detection project and start recognizing sign language gestures in real-time today !

‍
Nanogpt - GPT from scratch

Introducing the Generatively pre trained Transformer - a cutting-edge deep learning project that builds on the ideas presented in "Attention is All You Need" and OpenAI's GPT-2 / GPT-3 to create a powerful language model that can generate human-like text.

This project uses an auto regressive language modeling framework and PyTorch nn to train a GPT model, which can generate text by predicting the next word in a sequence based on the previous words. With its advanced architecture and powerful training algorithms, the GPT model can produce highly coherent and relevant text, making it an ideal tool for a wide range of applications, including natural language processing, content generation, and chatbots.

The Generatively Pretrained Transformer project builds on the success of ChatGPT and GitHub Copilot, which are both based on the GPT architecture.

So, if you're interested in exploring the exciting world of deep learning and natural language processing, head over to the Generatively Pretrained Transformer project and start generating text like never before !

Exploratory Analysis of Geolocational Data

Introducing the Exploratory Analysis of Geolocational Data - a powerful data analysis project that uses K Means Clustering Algorithm on geolocational data to recommend accommodations to immigrants in Bangalore city.

This project analyzes data on the geolocational distribution of various accommodations in Bangalore city, including apartments, hotels, and hostels. By applying K Means Clustering Algorithm, the project identifies clusters of similar accommodations and recommends the most suitable options to immigrants based on their preferences and budget.

The project also includes interactive Folium maps that show the geolocational distribution of accommodations in Bangalore city. While these maps may not be rendered on GitHub, you can view them through the provided link to get a better understanding of the geolocational analysis.

So, head over to the Exploratory Analysis of Geolocational Data project and start exploring the geolocational data of Bangalore city today !

Case Study Cyclistic Divvy Bikes

Introducing the Cyclistic Divvy Bikes case study project - a comprehensive data analysis project completed by Subhradip Roy as the Capstone Project of Google Data Analytics Professional Certificate.

This project follows the data analysis process of Ask, Prepare, Process, Analyse, Share, and Act, using R and Git/GitHub to host the resulting website. By analyzing the historical trip dataset of Cyclistic, a fictional bike-sharing company, the project answers key business questions about user behavior and usage patterns.

The datasets used in the case study have been provided by Motivate International Inc. and are appropriate for answering the business questions. The resulting website demonstrates the data analysis process and provides insights into the user behavior of Cyclistic's bike-sharing service.

So, head over to the project repository and check out the resulting website to see the insights and analysis for yourself !

Check out all my projects.