Sudipta Sarkar

Sudipta Sarkar

PhD Researcher (Graduate Student) · Computational & Data Sciences · Vision, Language & Retrieval
IISER Kolkata Kolkata, India Advisor: Dr. Dwaipayan Roy

About

I am a PhD scholar in the Department of Computational and Data Sciences at IISER Kolkata, where I am advised by Prof. Dwaipayan Roy. My research interests broadly encompass learning and understanding across vision, language, and information retrieval, with a particular focus on developing efficient and reliable methods for multimodal retrieval and reasoning, alongside research on resource-efficient visual and video understanding, representation learning, and adaptation.

Prior to joining IISER Kolkata, I worked as a Junior Research Fellow at IIT Kharagpur under the supervision of Prof. Abir Das. My research focused on resource-efficient video action recognition, self-supervised learning, long-context understanding, and video representation learning. I received my M.Sc. in Computer Science from Ramakrishna Mission Residential College, Narendrapur and my B.Sc. in Computer Science from Ramakrishna Vivekananda Centenary College, Rahara.

Recent Activities

Research Interests

Vision & Video Understanding Representation Learning Efficient Long-Context Understanding Information Retrieval Bibliometrics & Scholarly Analytics Domain Adaptation Language & Text Processing Information-Theoretic Learning Generative Modeling Medical Imaging Robotics & Automation

About

Academic background and personal journey

Biography

I am Sudipta Sarkar (Rik), a computer science researcher with a deep interest in learning, understanding, and the broader pursuit of knowledge. I am also deeply inspired by the teachings and ideals of Swami Vivekananda, Sri Ramakrishna, and Holy Mother Sri Sarada Devi.

I am currently pursuing my PhD in the Department of Computational and Data Sciences (CDS) at IISER Kolkata, where I am advised by Prof. Dwaipayan Roy.

Prior to joining IISER Kolkata, I worked as a Junior Research Fellow (JRF) in the Department of Computer Science and Engineering (CSE) at IIT Kharagpur, under the supervision of Prof. Abir Das.

I completed my M.Sc. in Computer Science in June 2025 from Ramakrishna Mission Residential College (Autonomous), Narendrapur, and my B.Sc. in Computer Science (Hons.) from Ramakrishna Mission Vivekananda Centenary College, Rahara. I also qualified the UGC NET examination in the June 2025 cycle for Assistant Professorship and PhD eligibility.

My academic journey has been closely connected with the Ramakrishna Mission, through which I had the opportunity to visit several of its centres and interact with teachers, monks, and fellow students. Their teachings and values have influenced my approach to learning, discipline, service, and personal growth. I remain deeply grateful to my family, teachers, friends, and mentors whose guidance has shaped both my academic and personal journey.

The name “Sudipta” (সুদীপ্ত) has Sanskrit origins, combining “Su”, meaning good, auspicious, or excellent, and “Dipta”, meaning bright, shining, or illuminated. It is therefore commonly interpreted as “bright” or “one who is illuminated.”

Skills

Python C/C++ Data Structures & Algorithms Database Systems Machine Learning Deep Learning Generative Models Large Language Models Information Retrieval Natural Language Processing Computer Vision Representation Learning Domain Adaptation Theoretical Computer Science PyTorch LaTeX Git SLURM

Languages & Interests

Languages

  • Bengali – Native Proficiency
  • English – Professional Working Proficiency
  • Hindi – Elementary Proficiency

Interests

Research

Projects, interests, and ongoing work

Research Interests

View My Research Statement

Vision & Video Understanding Representation Learning Efficient Long-Context Understanding Information Retrieval Bibliometrics & Scholarly Analytics Domain Adaptation Language & Text Processing Information-Theoretic Learning Generative Modeling Medical Imaging Robotics & Automation

Projects

Resource-Efficient Learning for Video Scene Understanding (RLV)
JRF Project · IIT Kharagpur · 2025 – 2026
Developing lightweight spatiotemporal models for efficient video scene parsing and action recognition. Leveraging state-of-the-art architectures including DeiT, Reversible MViT, ATLAS, Swin Transformer, Hiera, and ViT baselines. Key contributions include integration of FlashAttention-2 for faster and memory-efficient training, knowledge distillation from large pretrained models, multimodal fusion (RGB + optical flow), and SIFAR-based video-to-image reformulation.
Tools: PyTorch, FlashAttention-2, einops, ViT, MViT, RevMViT, Swin, Hiera, DeiT, ATLAS, DINOv3, SIFAR
Supervisor: Prof. Abir Das, IIT Kharagpur
Details coming soon
GRTS: A Gated Residual Teacher–Student Framework
Independent Research · 2026
Proposed a feature-level teacher–student refinement framework where a frozen self-supervised DINOv2 teacher provides stable semantic priors and a trainable ResNet-50 student learns dataset-specific residuals via a learnable gated fusion mechanism. Achieved 97.8% Top-1 accuracy on UC Merced and 96.8% on AID aerial image benchmarks.
Tools: PyTorch, DINOv2, ResNet-50, Gated Residual Fusion
GitHub
Super Image for Efficient Large-Scale Video Action Recognition
M.Sc. Thesis · 2025 · Advisor: Prof. Abir Das
Innovated a frame-aggregation pipeline using the SIFAR framework to recast video action recognition as image classification. Frames are rearranged via einops into super-images (e.g., 3×3 grid of 224×224 frames → 672×672 or 4×4 grid → 896×896) and processed with ImageNet-21K pretrained Hiera-ViT and ViT models. Achieved ~79% accuracy on Kinetics-400 and ~64% on Something-Something V2.
Tools: PyTorch, Hiera Vision Transformer, ViT, SIFAR, einops
GitHub
Image Steganography and Steganalysis
RKMRC · 2024
Developed a deep learning framework for stego/non-stego image classification. Extracted texture features using Gray Level Co-occurrence Matrix (GLCM), followed by pretrained ResNet50 for feature encoding, LSTM for sequence modeling, and a classification head.
Tools: PyTorch, ResNet50, LSTM, GLCM
Supervisors: Dr. Siddhartha Banerjee & Bibek Ranjan Ghosh
Coming soon
Facial Expression Recognition
B.Sc. Final Year Project · 2023
Engineered a compact pretrained VGG16-based model for real-time detection of 7 facial emotions (anger, disgust, fear, happiness, sadness, surprise, neutral) on the FER-2013 dataset. Emphasized quantization and low-resource optimization for potential mobile HCI deployment.
Tools: TensorFlow, VGG16, CNN
Supervisors: Prof. Chayan Halder & Prof. Prasenjit Das
GitHub
Nuclei Segmentation using UNet
Summer Project · 2023
Adapted UNet architecture for efficient segmentation of cell nuclei in biomedical microscopy images and videos. Explored potential extensions such as UNet++ and TransUNet for improved accuracy in resource-constrained settings.
Tools: TensorFlow, UNet
Supervisor: Dr. Biswajit Biswas
GitHub
Potato Disease Classification using CNN
Completed Feb 2023
Classifies potato leaf diseases using deep learning to improve early detection and crop cultivation in agriculture.
Tools: Python, Deep Learning, CNN
GitHub
Virtual Assistant Using OpenAI
Completed Nov 2022
Developed a voice-interactive virtual assistant mimicking Amazon Alexa or Siri using the OpenAI API and Python libraries.
Tools: Python, OpenAI API, pyttsx3, speech_recognition, webbrowser
GitHub
Number Tracking Game Using Backtracking Algorithms
Completed May 2022
CLI-based game in C++ where players race to reach a target number using backtracking algorithms, designed for the machine to win if it selects first.
Tools: C++
GitHub
Data Structure and Algorithm - Open Source
Completed May 2022
Contributed to an open-source collection of data structure and algorithm implementations for B.Sc. Computer Science students per UGC syllabus.
Tools: C++, Data Structures, Algorithms
GitHub
Medical Store Billing Management System Software
Completed May 2021
CLI-based billing system for medical stores developed in C++ during my first semester.
Tools: C++
GitHub

Publications

Peer-reviewed papers and preprints

Conference Papers

  • Image Classifiers are Efficient Self-Supervised Video Representation Learners
    Owais Iqbal, Sudipta Sarkar, Shyam Marjit, Omprakash Chakraborty, Anirban Chakraborty, Abir Das
    The 37th British Machine Vision Conference (BMVC 2026) · Accepted
    Project Page PDF · Coming Soon PPT · Coming Soon Code · Coming Soon

Thesis

  • Super Image for Efficient Large-Scale Video Action Recognition
    Sudipta Sarkar
    M.Sc. Thesis · Ramakrishna Mission Residential College (Autonomous), Narendrapur · 2025
    PDF · Coming Soon Slides

Teaching & Talks

Presentations, seminars, and mentorship

Talks & Presentations

  • Resource Efficient Visual Understanding
    PeRCeiVe Lab, University of Catania · June 16, 2026 · Remote

    Presented research on resource-efficient visual understanding, domain adaptation and discussing approaches for developing computationally efficient and scalable learning systems.

    Presentation Slides · Coming Soon

  • Resource-Efficient Large-Scale Video Action Recognition Using Image Classifiers
    CVIR Group Reading Room · IIT Kharagpur · December 22, 2025

    Presented research on resource-efficient large-scale video action recognition using image classifiers, with a focus on super-image-based video representation learning.

    Venue: CSE 126, Department of Computer Science and Engineering, IIT Kharagpur

    Presentation Slides · Coming Soon

  • Seminar on Image Generation and Diffusion Models
    RKMRC, Narendrapur · May 17, 2025

    Delivered a seminar covering the foundations of diffusion models and their applications in image generation.

    View Seminar Slides

  • Final Semester Thesis Presentation
    RKMRC, Narendrapur · May 19, 2025

    Presented the M.Sc. thesis, “Super Image for Efficient Large-Scale Video Action Recognition.”

    View Thesis Presentation

  • Technical Articles & Writings
    Ongoing

    Technical writings covering topics in Complexity Theory, Linear Algebra, Backpropagation, and mathematical proof methodologies.

    Explore Technical Articles

Teaching & Mentorship

  • Teaching Assistant — CS3102
    IISER Kolkata · August 2026 – December 2026 · Ongoing

    Serving as a Teaching Assistant for CS3102 during the Autumn semester at the Department of Computational and Data Sciences, IISER Kolkata.

    Course Instructor: Dr. Dwaipayan Roy

  • B.Tech. Thesis Mentorship
    IIT Kharagpur · December 2025 – March 2026

    Mentored a B.Tech. student at IIT Kharagpur on their undergraduate thesis on fine-grained image classification in long-context images.

    Primary Supervisor: Prof. Abir Das

  • Co-Organizer, Neuroverse Coding Competition
    RKM Vivekananda Centenary College, Rahara · March – April 2023

    Designed and curated programming problem sets for Neuroverse, a college-level coding competition.

  • IT Sub-Committee Member, Vidyarthi Sabha
    RKMRC, Narendrapur · September 2023 – September 2024

    Contributed to the Vidyarthi Sabha IT Sub-Committee by providing technical support and IT consulting for institutional activities.

Education & Experience

Academic background and professional experience

Education

  • Indian Institute of Science Education and Research (IISER) Kolkata
    Ph.D. in Computational and Data Sciences · August 2026 – Present

    Advisor: Prof. Dwaipayan Roy, Assistant Professor, Department of Computational and Data Sciences, IISER Kolkata.

    Research Area: Learning and Understanding Across Vision, Language, and Information Retrieval

  • Ramakrishna Mission Residential College (Autonomous), Narendrapur
    M.Sc. in Computer Science · September 2023 – June 2025

    Thesis: Super Image for Efficient Large-Scale Video Action Recognition

    Advisor: Prof. Abir Das, Associate Professor, Department of Computer Science and Engineering, IIT Kharagpur.

    Department Rank: 1st (Rank 1)  ·  CGPA: 9.95/10.00 (96.30%)  ·  Approx. GPA: 4.0/4.0 (U.S. scale)

    European Equivalent: approximately A / 1.0–1.3 (country- and institution-dependent)

  • Ramakrishna Mission Vivekananda Centenary College (Autonomous), Rahara
    B.Sc. (Hons.) in Computer Science · September 2020 – May 2023

    Thesis: Human Facial Expression Recognition

    Advisors: Prof. Chayan Halder and Prof. Prasenjit Das.

    Department Rank: 3rd (Rank 3)  ·  CGPA: 9.72/10.00 (92.02%)  ·  Approx. GPA: 3.9–4.0/4.0 (U.S. scale)

    European Equivalent: approximately A / 1.0–1.5 (country- and institution-dependent)

  • Hogalbaria Adarsha Siksha Niketan (H.S.), Hogalbaria, Nadia
    Higher Secondary · May 2018 – May 2020

    Subjects: Physics, Chemistry, Mathematics, Biology  ·  Percentage: 83.00%

    Board: West Bengal Council of Higher Secondary Education (WBCHSE)

Experience

  • Indian Institute of Technology Kharagpur
    Junior Research Fellow · November 2025 – March 2026

    Department: Computer Science and Engineering

    Project: Resource-Efficient Learning for Video Scene Understanding (RLV)

    Supervisor: Prof. Abir Das

  • Indian Institute of Technology Kharagpur
    Research Intern · January 2025 – October 2025

    Department: Computer Science and Engineering

    Project: Resource-Efficient Learning for Video Scene Understanding (RLV)

    Supervisor: Prof. Abir Das

  • Ramakrishna Mission Residential College (Autonomous), Narendrapur
    IT Sub-Committee Member, Vidyarthi Sabha · September 2023 – September 2024

    Contributed to the IT Sub-Committee of Vidyarthi Sabha, providing technical support, IT consulting, and assistance with technology-related activities.

  • Ramakrishna Mission Vivekananda Centenary College, Rahara
    Co-Organizer, Neuroverse Coding Competition · March 2023 – April 2023

    Co-organized the college-level Neuroverse coding competition and contributed to the design and curation of programming problem sets.

Achievements

  • Best Postgraduate Student Award
    Academic Year 2023–2025 · December 2025

    Awarded by Ramakrishna Mission Residential College for outstanding academic performance during the M.Sc. programme.

  • UGC NET Qualified — 99.83 Percentile
    June 2025 Cycle · July 2025

    Qualified the UGC NET examination for Assistant Professorship and PhD eligibility.

  • 1st Rank — M.Sc. in Computer Science
    June 2025

    Graduated with Department Rank 1 and a CGPA of 9.95/10.00 (96.30%).

  • National Scholarship for Post Graduate Studies
    Academic Year 2023–2025 · October 2024

    Awarded under the national postgraduate scholarship programme, which supports high-achieving students across India.

  • 3rd Rank — RKMVERI M.Sc. Admission Test
    June 2023

    Secured 3rd rank in the M.Sc. admission test conducted by Ramakrishna Mission Vivekananda Educational and Research Institute.

  • 3rd Rank — B.Sc. in Computer Science
    May 2023

    Graduated with Department Rank 3 and a CGPA of 9.72/10.00 (92.02%).

  • 1st Rank — Intra-College Coding Competition
    May 2022

    Secured 1st position in an intra-college programming competition.

Curriculum Vitae

Access my detailed Curriculum Vitae for a complete record.

View CV on Google Drive

Download My Curriculum Vitae (CV)

Contact

Get in touch

Contact Information

X (Twitter)
@sarkar360000
YouTube
Rik Sarkar
Current Address
Research Scholar Lab (Room No. 106), First Floor,
Department of Computational and Data Sciences (CDS),
IISER Kolkata, Mohanpur, Nadia,
West Bengal 741246, India
Old Address
Takshashila Building (2nd Floor, Research Scholar Lab),
IIT Kharagpur, Kharagpur, West Bengal 721302

Academic References

  • Dr. Dwaipayan Roy

    Assistant Professor, Department of Computational and Data Sciences, IISER Kolkata. PhD Advisor.

    Email: dwaipayan.roy@iiserkol.ac.in

    Faculty Profile

  • Dr. Abir Das

    Associate Professor, Department of Computer Science and Engineering, IIT Kharagpur. M.Sc. Thesis Advisor.

    Email: abir@cse.iitkgp.ac.in

    Faculty Profile

  • Dr. Chayan Halder

    Assistant Professor, Department of Computer Science, RKM Vivekananda Centenary College, Rahara. B.Sc. Thesis Advisor.

    Email: chayan.comp@rkmvccrahara.org

  • Dr. Siddhartha Banerjee

    Associate Professor and Head of Department, Ramakrishna Mission Residential College (Autonomous), Narendrapur. M.Sc. Course Instructor.

    Email: sidd_01_02@yahoo.com

  • Swami Iswarananda

    Vice Principal, Ramakrishna Mission Residential College (Autonomous), Narendrapur. M.Sc. Academic Mentor.

    Email: vp@rkmrc.in

  • Swami Ekachittananda

    Principal, Ramakrishna Mission Residential College (Autonomous), Narendrapur. M.Sc. Academic Mentor.

    Email: rkmcnpur@gmail.com

  • Swami Vedanuragananda

    Vice Principal, Ramakrishna Mission Vivekananda Centenary College, Rahara. B.Sc. Academic Mentor.

    Email: bodhisattamj@gmail.com