Real Time Crime Detection
The proposed approach extracts frame-level characteristics using convolution and LSTM layers to capture the long-term relationships of activities in surveillance videos.
Demo
Project Trinetra
Project Trinetra is an early-stage research project inspired by Google's Guideline project. Its primary objective is to enable individuals with low vision to run independently using computer vision and machine learning. The core technology utilized in Trinetra is a semantic segmentation model that classifies every pixel in the frame as "walkable" or "not walkable" to predict the runner's position relative to the safe path.
Demo
Instacart Product Recommendation
The Instacart Product Recommendation project aims to enhance the user experience by providing accurate and relevant product recommendations based on users' past order history. By building a machine learning model that predicts which products are likely to be reordered, Instacart can simplify the process of ordering groceries online and increase user satisfaction.
Demo
Autism Screening using Machine Learning
The project aims to improve autism screening by creating a machine learning model that predicts the likelihood of an individual having autism spectrum disorder (ASD) based on survey data. The machine learning model developed in this project helps healthcare professionals prioritize their resources by identifying individuals who may require further evaluation for ASD.
Demo