Anirban Mukhopadhyay

Anirban Mukhopadhyay

Ph.D. student, Department of Computer Science

Virginia Tech

Biography

I am a second-year Ph.D. student in the Department of Computer Science at Virginia Tech, where I am advised by Dr. Kurt Luther in the Crowd Intelligence Lab . Prior to joining grad school, I was a Software Engineer at Microsoft for two and a half years. I obtained a Bachelor of Engineering degree in Computer Science and Engineering from Jadavpur University, India.

In my research, I take a mixed-methods approach that combines my skills in software development and human-computer interaction. The work entails the design and evaluation of sociotechnical systems that support expert-led crowdsourcing and collaboration. Overall, I am interested in human-AI interaction projects that aim to augment the work practices of experts.

Download my resumé

Interests
  • Human-AI Interaction
  • Crowdsourcing
  • Future of Work
Education

  • Ph.D. in Computer Science, 2021 - Present

    Virginia Tech

  • M.S. in Computer Science, 2021 - Present

    Virginia Tech

  • Bachelor of Engineering in Computer Science and Engineering, 2014 - 2018

    Jadavpur University, India

Skills

Full-stack Development

REST APIs, SQLServer, PostgreSQL, DynamoDB, Django, React, Heroku app hosting

Data analysis

Pandas, Numpy, Scikit-Learn, Seaborn, Matplotlib

Design Research

Survey Design, Semi-structured interview and Thematic Analysis, Wireframing, Contextual Analysis

Experience

 
 
 
 
 
Crowd Intelligence Lab, Virginia Tech
Graduate Research Student
Crowd Intelligence Lab, Virginia Tech
Jan 2021 – Present Blacksburg, Virginia
  • Research Assistant at the Crowd Intelligence Lab with Dr. Kurt Luther.
  • Relevant Coursework - Human-AI Interaction, Data Analytics, Deep Learning, Usability Engineering, Computer Supported Collaborative
 
 
 
 
 
Microsoft
Software Engineer
Microsoft
Jun 2018 – Dec 2020 Hyderabad, India

Responsibilities included:

  • Full stack development using C#, React, Python
  • Improvement of service reliability and API performance
  • Proof of concept for auto-tagging feature in SharePoint
 
 
 
 
 
Amazon
Software Development Engineer Intern
Amazon
May 2017 – Jul 2027 Hyderabad, India
Created a migration tool to move data from relational database (Aurora DB) to NoSQL database (Dynamo DB)

Projects

Expert-led crowdsourcing in Open Source Intelligence (OSINT) Investigations

Developed a framework for students trained in social media information discovery and verification skills to augment the investigative practices of journalists, law enforcement officers, and human rights activists. I ran five study sessions where students contributed to real-world investigations being carried out by the experts.

I used design-based research (DBR) approach to develop the collaborative framework. Data collection was in the form of collaborative spreadsheets containing feedback from experts, semi-structured interviews with students and experts, and student reflection assignments.

The work is in preparation for CHI 2023.
Anirban Mukhopadhyay, Kurt Luther
Expert-led crowdsourcing in Open Source Intelligence (OSINT) Investigations
Collaboration and competition in Open Source Intelligence (OSINT) analysis

Developed a web app for students to participate in Capture the Flag (CTF) events to discover, archive, verify and report possible social media misinformation. The Quriosinty app was used by 40 students every week during the Spring ‘21 semester.

Here users can form teams, submit pieces of investigation through flags and judges can award points based on a rubric. I built REST API endpoints to extend the platform for integrating new OSINT tools.

We used a Research through Design (RtD) process to iterate on the platform and generate design recommendations for systems aimed at verification of online information. Data was collected from system logs, student reflections, and semi-structured interviews.

The paper will be submitted to CSCW 2022.
Sukrit Venkatagiri, Anirban Mukhopadhyay, Kurt Luther
Collaboration and competition in Open Source Intelligence (OSINT) analysis
Explorable Interactive Human Reposing

Human-AI Interaction Course project 2021

We design a notebook to allow the user to upload an image of a person and modify the pose of that person by dragging and dropping body joints.

We use CoCosNet-v2 to synthesize the image of the reposed person. We use OpenPose to extract the pose of the person. We use IPython widgets to enable interaction with the extracted pose.

Anirban Mukhopadhyay, Badour AlBahar, Hongjie Chen
Explorable Interactive Human Reposing
Red Flags: Combating ad hominem attacks in comment-based online discussions

Computer Supported Collaborative Work Course project 2021

Here we propose a flagging mechanism that modifies the process of reporting currently found in popular platforms like Facebook, Instagram, and Twitter. We describe the design, development, and evaluation of a method of curbing ad hominem attacks that occur in common online discussion spaces, particularly comments and replies to posts. Through ethnographic interviews as well as a small-scale user study, we explore the emotional experiences of being the target of or witnessing trolling online, revealing the needs of these groups in these interactions.
Anirban Mukhopadhyay, Juliet Clarke, Liling Yuan
Red Flags: Combating ad hominem attacks in comment-based online discussions
Study and improvement of accuracy in the handwritten Indian script recognition problem
Used texture and frequency-based features along with CNN. Performed classifier combination to achieve high accuracy in the script identification task.

Papers published:
[1] Mukhopadhyay, Anirban, et al. “A study of different classifier combination approaches for handwritten Indic Script Recognition.” Journal of Imaging (2018): 39.
[2] Mukhopadhyay, Anirban, et al. “Handwritten Indic script recognition based on the Dempster–Shafer theory of evidence.” Journal of Intelligent Systems 29.1 (2020): 264-282.
Anirban Mukhopadhyay, Ram Sarkar
Study and improvement of accuracy in the handwritten Indian script recognition problem

Contact