Shekhor Chanda

I completed my master's in Computer Science from the University of Manitoba. In my master’s research, I was working with different Computer Vision problems using Deep Learning, Machine Learning techniques such as Domain Adaptation, Knowledge Distillation, Model Adaptation, Transfer Learning, etc. I also have strong knowledge in Machine Learning Algorithms such as SVM, Linear Regression, Logistic Regression, Decision Trees, and different frameworks like Pytorch, Keras, etc.

Before doing my master's, I worked as a Software Engineer for two and half years. Where most of the projects I was working with JAVA-based framework Spring and Spring-Boot. Also, I have some experience in C++ development. I hope this experience will help me to contribute to the team. Besides these, I have a strong background in problem-solving with Data structure and Algorithms.

I am looking for an opportunity to prove my skills and knowledge to real-world problems.

Graduate Research Assistant

• Black-box model adaptation and distillation: We adopt a Deep Learning model for recognizing objects in the target domain without accessing any internal information of the source model. We use the side-tuning and Hypothesis Transfer technique for the adaptation and then we consider a Knowledge Distillation technique to build a compact white-box model. We also apply Pseudo Labeling idea during the distillation.
• Dynamic Transfer for domain adaptation in Crowd Counting: A deep learning model is trained dynamically for crowd counting. The core idea of dynamic transfer is to adapt the model to the target domain by dynamically adapting it to individual data samples.

Software Engineer

• Designing the APIs using JAVA, Spring, REST, Soap API, SQL and integrating with front-end clients using REST client, JSP, HTML, CSS, Javascript, JQuery to develop fully automated Web applications.
• Worked on key verification tasks using open source C++ project SIGNSERVER and HSM, JBOSS.
• Implemented an internal messaging service using RabbitMQ message broker.

Master's in Computer Science 2021-2022

Bachelor in Computer Science 2014-2018