Adagrad AI
Machine Learning Engineer, working on cutting edge solutions in the field of computer vision for real-time video analytics and access-control through applications using Person Detection, ReID and Automated License Plate Recognition.
I'm an engineer with passion for data driven problem solving. Looking to collaborate with people from diverse backgrounds and expertise to learn, grow and provide value to society.
Simple and effective tutorials on getting started with your journey in Computer Vision.
Machine Learning Engineer, working on cutting edge solutions in the field of computer vision for real-time video analytics and access-control through applications using Person Detection, ReID and Automated License Plate Recognition.
Machine Learning Engineer, developing Computer Vision Projects on domains such instance segmentation and semantic segmentation for visual distance measurement. Productionizing and optimizing pipelines to achieve efficient inference.
Interned and worked on a project focused at providing a natural language interface for databases. Applied NLP concepts in domains such as Entity Extraction and Intent Recognition for converting natural language queries into database queries.
A content based fashion products recommendation Engine which combines various aspects of Computer vision to perform object detection and image embedding generation. The image embeddings are used to recommend similar products from a catalog of fashion items. Inspired by a research article from Myntra.
A dataset has been prepared by the engineers in Daimler to estimate and reduce the bench time for testing automobiles, it contains combinations of car features, configurations and different testing methods applied with their respective time required for testing. Goal is to model the time taken by each interaction.
An in-depth analysis aimed at suggesting suitable locations to set up a restaurant as well as the type of restaurant based on the characteristics of the location. The suggested locations are selected, keeping in mind the possibilities for future expansion.
Neural Network / Multi Layered Perceptron implemeted from scratch using just vector and matrix computations. The goal of the projects was to understand the intricacies of Deep Neural Network training, gradient flow during backpropagation of loss and experimenting with various optimization methods for faster convergence.