Work Experience

ARTIFICIAL INTELLIGENCE ENGINEER

1st May,20 - Present

Actively Participated in Product Management, Competitor Analysis as a part of CTO Team.
Worked on RZT Document Processor for LIBOR AI \& Document Question Answering System.
Built Cloth Semantic Segmentation for Apparel Similarity Module by using U-net.
Implemented ‘Integrated Gradient’ feature as a Marketplace Block for various tasks (Text, Image Classification etc).
Big Data Processing of 4 million records by using PySpark for Fraud Customer Detection.
Built Anomaly Detection in Mechanical Parts by using Multi Layer Feature Sparse Encoding & Autoencoder.

DEEP LEARNING RESEARCHER

1st September,19 - 30th April,20

Working on Domain Adaptive Semi-Supervised Classification Problem for Sparrosense AI Supervisor Product.
Worked on Dialogue Act Classification from Telephone Calls.
Worked on Human-Object Interaction Classification Problem.
Worked on Semantic Segmentation of Steel Surface Defects from Steel Sheets.

UNDERGRADUATE RESEARCH ASSISTANT

8th Oct,17 – 28th Feb,19

Worked on Computer Vision, 3D Point Cloud Processing Research Problems for Developing a Vision Based Control System of Quadruped Robot under joint guidance of Dr. Ananda Shankar Chowdhury & Dr. Sanjoy Kumar Saha.
Created Synthetic Dataset of LIDAR Data for Our Project by using Blensor with Velodyne LIDAR.

UNDERGRADUATE RESEARCH ASSISTANT

1st March,17 – 30th Sept,17

Worked on Research Problem on Devloping Language Indentification , Sentiment Analysis Model for Code-Mixed Data under guidance of Dr. Dipankar Das.
Created a Bengali-English Code-mixed Dataset for NLP tools contests at ICON 2017.

Internships

CONCURRENT TEMPORAL ACTION RECOGNITION USING INFLATED 3D CONVNETS (I3D)

7th June,19 - 31st August,19

Research Internship at “Video Analytics Lab” at Silversparro for research & developing the solution for “ Fast Concurrent Temporal Action Recognition”.
Contributed various Object Detection pipelines for ’Multiview Person Detection’ using Deformable-ConvNet(DCN) , RetinaNet etc.

DYNAMIC OBJECT EXTRACTION FROM MOVING MONOCULAR CAMERA

18th May,18 - 30th July,18

Summer Internship at “Autonomous Drone Lab of TCS Innovation Lab , Kolkata” under Dr. Arnab Sinha. for research & developing the solution for “Dynamic Object Extraction from Moving Monocular Camera” project which is a part of MBZIRC 2020 Challenge.
Implemented a ROS packages for Automated Photometric Calibration.
Done a comparative study between different types of Monocular SLAM Algorithms (like ORB SLAM ,DSO) under different conditions like variable intensity , dynamic environment etc.

ADVANCED LANE DETECTION IN THE INDIAN ROAD DURING DAY & NIGHT TIME USING DEEP LEARNING

16th Dec,17- 20th Feb,18

This project is a virtual internship at a startup “Giscle” for “Designing and Development of algorithm & it’s module for Lane Detection in the Indian road during day & night time using Deep Learning. This module is a part of their kit for their Product for Autonomous Cars.
implemented the lane detection for both straight & curved lanes using regression & RANSAC algorithm.

DEVELOPMENT OF DATA EMBEDDING AND RETRIEVAL SYSTEM FOR IMAGE ,AUDIO ,VIDEO FILES USING INVISIBLE WATERMARKS

27th Augt,17 - 8th Oct,17

This project is a virtual internship at a startup “NAXAP” for “ Designing and Development of Data embedding and retrieval system for image ,audio ,video files using invisible watermarks” for their product. This extension is made to protect the copyright and ownership of the files/document and share it more securely.
Implemented both lossy & lossless invisible watermarking techniques based on their processing time.

AUTO-SUMMARY GENERATOR FOR A TOPIC IN A SEARCH ENGINE

20th Jun,17 - 20th Aug,17

Summer Internship at “DataWUDI” for “Designing and Development of a NLP & AI based Web Application on Auto-Summary Generator for a topic in a search engine”.
Contributed to Auto-summarization & DL based context detection part of the application.