About Me
-
Worked on Object Detection and Recognition for product recognition in Smart Store using Deep Learning Algorithms.
-
Worked on Human Tracking for person re-identification using Machine Learning and Deep Learning Algor...
-
Developed an annotation Tool using PyQt5 GUI, used for product selection.
-
Developed an Active Learning Tool based on Interpolation concept and is used for synthetic annotation for product bounding boxes in an images, which reduces the annotation teamwork.
-
Working on Measuring Anthropometric measurements from a Human body obj file (3D).
-
Working on Video-based reconstruction of 3d people models.
-
Working on segmenting the 3d Human mesh and Cloth wrapping on the human mesh.
-
Got an Experience on Understanding Deep Learning models from scratch and how deep learning can be applied on Object -Detection, Object Tracking and Person re-identification.
-
Gained Knowledge of Machine Learning and Deep Learning Algorthims.
-
Acquired Experience of Data handling methods like data correction, data augmentation, making correct data before training the model.
-
worked on the deployment the deep learning or machine learning models using Flask.
Skills
Web Development
Programming Language
Database
Operating System
Data & Analytics
Others
Positions
Portfolio Projects
Company
JoMart
Role
Data Scientist
Description
Jo-Mart describes JoMart as "a new kind of store with no checkout required". That means, when you shop at JoMart, you'll never have to wait in line. The store works with a Jo-Mart application - you enter JoMart, take the products you want and, thanks to the app, just leave again.ds
Show More Show LessTools
Sublime TextCompany
JoSmartstore
Role
Data Scientist
Description
JoSmartstore’s cutting edge AI solution creates on screen 3D avatar of customer wearing the selected dresses and accessories creating a trial room experience virtually. 3D models of products and accessories which the store sells are draped on to the 3D avatar of the customer displayed on screen. It enables consumers to checkout the selected products and accessory on their virtually created avatars. The customers avoid long queues, fitting challenges and the hassle of physically changing into multiple dresses. They can select items from the on screen menu and try on many dresses and accessories to their complete satisfaction without going back and forth.
Show More Show LessSkills
Keras TensorFlow NumPy PyMeshTools
Sublime Text Blender Unity 3D