Aman Shrivastava

Aman Shrivastava

CS PhD Student

I am a Computer Science PhD Candidate at the University of Virginia advised by Prof. Tom Fletcher and previously by Prof. Vicente Ordóñez Román. During my PhD, I interned at Adobe Reserach where I worked with Kushal Kafle on bootstrapping large language and vision models to design VLMs via instruction-tuning. I also interned with Salesforce Research where I worked with Nikhil Naik and Prof. Stefano Ermon on vision-language alignment & retrieval and conditional & controllable image generation using diffusion models. Previously, I worked as a Reserach Scientist with Prof. Donald E. Brown where I developed AI methods for disease understanding and diagnosis. I also hold a Masters in Data Science from the University of Virginia and a Bachelors in Technology from the Indian Institute of Technology, Roorkee.

My research areas include (1) vision-language alignment and representation learning, (2) conditional and controllable image and text generation, (3) foundation language and vision models, as well as (4) their applications for disease understanding and addressing challenges in computational medical imaging. Find my CV here.

📝 Selected Publications

NASDM, Nuclei-Aware Semantic Histopathology Image Generation Using Diffusion Models
Aman Shrivastava and P. Thomas Fletcher
In International Conference on Medical Image Computing and Computer-Assisted Intervention. MICCAI 2023 (Oral)
paper | code

CLIP-Lite, Information Efficient Visual Representation Learning from Textual Annotations
Aman Shrivastava, Ramprasaath R. Selvaraju, Nikhil Naik, and Vicente Ordonez.
In International Conference on Artificial Intelligence and Statistics, PMLR. AISTATS 2023 (Oral)
paper | code

Estimating and Maximizing Mutual Information for Knowledge Distillation
Aman Shrivastava, Yanjun Qi, and Vicente Ordonez.
In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. CVPR TCV Workshop 2023
paper | code

Improving interpretability via explicit word interaction graph layer
Arshdeep Sekhon, Hanjie Chen, Aman Shrivastava, Zhe Wang, Yangfeng Ji, and Yanjun Qi.
In Proceedings of the AAAI Conference on Artificial Intelligence. AAAI 2023
paper | code

Identifying metabolic shifts in Crohn’s disease using’omics-driven contextualized computational metabolic network models
Philip Fernandes, Yash Sharma, Fatima Zulqarnain, Brooklyn McGrew, Aman Shrivastava, Lubaina Ehsan, Dawson Payne et al.
In Nature Scientific Reports 2023.
paper

Self-attentive adversarial stain normalization
Aman Shrivastava, William Adorno, Yash Sharma, Lubaina Ehsan, S. Asad Ali, Sean R. Moore, Beatrice Amadi, Paul Kelly, Sana Syed, and Donald E. Brown.
In International Conference on Pattern Recognition, Springer. ICPR 2021 (Oral)
paper | code

Cluster-to-Conquer, A Framework for End-to-End Multi-Instance Learning for Whole Slide Image Classification
Yash Sharma, Aman Shrivastava, Lubaina Ehsan, Christopher A. Moskaluk, Sana Syed, and Donald E. Brown.
In Medical Imaging with Deep Learning, PMLR. MIDL 2021
paper | code

Deep learning for visual recognition of environmental enteropathy and celiac disease
Aman Shrivastava, Karan Kant, Saurav Sengupta, Sung-Jun Kang, Marium Khan, S. Asad Ali, Sean R. Moore et al.
In IEEE EMBS International Conference on Biomedical & Health Informatics (BHI). BHI 2019
paper

📂 Side Projects

Krity
Co-founded an open audiobook platform that allows listeners to find audiobooks in diverse voices, and narrators to give voices to their favorite books. Have produced and published over 40 audiobooks.
website

Connect 4 AI
Developed a lightweight connect-4 game with a self-written pure-javascript bot using Minimax algorithm and Monte Carlo simulations.
website | code

Humorous Image Captioning System
Implemented a self-attentive encoder-decoder framework to generate humorous captions for images indistinguishable from human generated memes.
code

Soccer Squad Optimization
Designed a strategic football squad selection algorithm given budget, nationality (and/or club) and playing formation constraints based on self extracted FIFA dataset. Longstanding featured dataset on Kaggle.
code | dataset

📖 Education

  • 2020 - present: Computer Science PhD | University of Virginia
  • 2018 - 2019: Masters in Data Science | University of Virginia
  • 2013 - 2017: Bachelors in Technology | Indian Institute of Technology, Roorkee

💻 Experience

  • Research Scientist Intern at Adobe Research | June - November, 2023
    Worked on designing an AI assistant for visual reasoning via bootstrapping pretrained foundation models.

  • Research Scientist Intern at Salesforce Research | June - November, 2022
    Worked on conditional generative diffusion models for image synthesis and vision-language alignment.

  • Research Scientist at University of Virginia | June 2019 - June 2020
    Developed learning frameworks for the understanding and assisted diagnosis of gastrointestinal diseases.

  • Analyst at Citibank | June 2018 - June 2019
    Built a streamlined visualization platform with data-driven insights for the Chief Country Officer. However, would not recommend.

👨‍🏫 Teaching / Talks

  • 2023
    • Co-instructor for Geometry of Data | University of Virginia | lecture videos
    • Oral Presentation at MICCAI 2023
    • Invited Speaker for Research Speaker Series | PathAI
    • Teaching Assistant for Digital Signal Processing | University of Virginia
  • 2022
    • Teaching Assistant for Digital Signal Processing | University of Virginia
    • Teaching Assistant for Geometry of Data | University of Virginia
    • Teaching Assistant for Machine Learning | University of Virginia
    • Python Instructor for SOAR Scholars Program | University of Virginia
  • 2019
    • Python Instructor for Health Sciences Library | University of Virginia
    • Assistant Capstone Advisor for School of Data Science | University of Virginia
    • Invited Speaker for Applied Machine Learning Conference | Tom Tom Festival