Aditya Narendra

I am an independent researcher working towards building safe, reliable, and fair ML systems for healthcare applications.

Previously, I was fortunate to work with Dr. Leland K Werden at ETH Zürich, Dr. Min Xu at Carnegie Mellon University (CMU), and Dr. J. Sivaswamy at IIIT Hyderabad. I also worked as a Software Engineer at Tech Mahindra's AI Lab.

Feel free to reach out to discuss research ideas or potential collaborations.

Email | CV | LinkedIn | GitHub

Updates
Our paper on Class-Conditional Coverage got accepted to AAAI 2025.
Our paper on Feature Bias Mitgation got accepted to WIML @ NeurIPS 2024.
Our paper on Uncertainty Quantification for Cervical Cytology got accepted to MIDL 2024.
Awarded 1st Prize ($2500 USD) out of 1000 teams at 2022 Smart Odisha Hackathon.
Awarded 2nd Prize ($200 USD) out of 100 teams at 2022 Hugging Face Gradio NYC Hackathon.
Got accepted to DLMI 2022 Summer School at ETS Montreal.
Graduated from OUTR and my UG thesis nominated for Best Thesis Award 2021.
Experience
Associate Software Engineer | Center of Excellence-Artificial Intelligence
Aug 2022 - Nov 2024

Worked on a GNN-based accident detection system for a smart traffic solution for the Govt. of Odisha, cutting emergency response times by over 60%. Also designed an EHR application handling 100,000+ daily records and improved record retrieval by 42%.

Research Affiliate | Assisted Forest Regeneration lab
Advisor: Dr. Leland K Werden
Dec 2022 - Jan 2024

Built a sapling detection algorithm that detects over 300 tree species, for savannah and mangrove restoration projects. Finetuned a Llama2-13b model for a summarization platform with custom review tags for grey literature of regeneration practices on ASReview Lab.

Research Intern | Xu lab
Advisor: Prof. Min Xu
Aug 2022 - Sept 2023

Worked on a Contrastive Self-Supervised Learning (CSSL) approach for macromolecular structure classification from cryo-ET data with limited labels. Also contributed to an unsupervised multi-task learning framework for 3D subtomogram image alignment, clustering, and segmentation in cryo-ET environment.

Research Volunteer | Summer School on Computational Neuroscience
Advisor: Dr. José Biurrun Manresa
July 2017 - Aug 2023

Participated in the 2023 Neuromatch Academy Summer School on Computational Neuroscience. Designed regression models for future motion state prediction using time series analysis on ECoG data.

Research Intern | Summer School on Deep Learning for Medical Imaging
Advisors: Prof. Pierre-Marc Jodoin & Prof. Thomas Grenier
Jul 2022 - Aug 2022

Partcipated in the 3rd Edition Summer School on Deep Learning for Medical Imaging (DLMI-22). Evaluated various weakly supervised segmentation techniques for cardiac diseases diagnosis.

Research Assistant | IHub-Data
Advisors: Prof. Jayanthi Sivaswamy & Prof. C.V. Jawahar
Jul 2021 - Jan 2022

Worked on building the 'Indian Brain Segmentation Dataset'- a sub-cortical structure segmentation database for young population. Also contributed to a multi-scale attention architecture for COVID-19 detection from Chest-X Rays.

Publications
Optimizing Conformal Prediction Sets for Pathological Image Classification
Shubham Ojha*, Aditya Narendra*, Abhay Kshirsagar, Shyam Sundar Debsarkar & Surya Prasath
Pattern Recognition (Under Review), 2025
Paper | Code
CP Training method for controlling set compostionality.
Ensuring Class-Conditional Coverage for Pathological Workflows
Siddharth Narendra, Shubham Ojha, Aditya Narendra, Abhay Kshirsagar & Abhisek Mallick
AAAI Conference on Artificial Intelligence (AAAI), 2025
Paper | Webpage | Code | Poster | Slides
CP method for consistent coverage guarantee across all classes.
Mitigating Feature Bias in DL Models for Cervical Cytology
Subhashree Sahu, Shubham Ojha & Aditya Narendra
WiML, Neural Information Processing Systems (NeurIPS), 2024
Paper | Webpage | Code | Poster | Slides
Sampling based technique for feature bias mitigation.
Uncertainty Quantification in DL Models for Cervical Cytology
Shubham Ojha & Aditya Narendra
Medical Imaging with Deep Learning (MIDL), 2024
Paper | Webpage | Code | Poster | Slides
Effect of uncertainty incorporation on model's predictive capabilities.
Deep Learning Based Classification of the Big Four Snake Species Using Visual Features
Nishikanta Parida, Aditya Narendra, Pooja Reddy Kolimi, Priyansu Panda & Ipsit Misra
IEEE Confluence, 2023
Paper | Slides
Transformer based model for Big-4 snake species classification.
From Robots to Books: An Introduction to Smart Applications of AI in Education (AIEd)
Shubham Ojha, Siddharth Mohapatra, Aditya Narendra & Ipsit Misra
Springer ICRIC, 2023
Paper | Slides
Survey on AI-driven applications in Education
Other Projects
Prediction of Future Continuous Motion States from ECoG Recording
Aditya Narendra, Andy Bonnetto, Chayanon Kitkana, Paola Juárez, Ruman Ahmed Shaikh & Taima Crean
2023 NMA Summer School on Computational Neuroscience
Slides | Code
Regression models for future motion state prediction on ECoG data.
MoSwasthya: ML Based Application for Cardiac Disease Risk Prediction
Aditya Narendra, Nishikanta Parida, S.R. Mohanty & Shaktee Biswal
2022 Smart Odisha Hackathon (1st Prize Winner)
Slides | Code | Video
Ensemble method-based estimation cardiac risk estimation.
Vision-Based Models for Sorting and Segregation of Waste
Omdena Community Project
Slides | Code
VGG 16-based model for waste materials sorting & segregation.
Miscellaneous
  • Taught (401-Deep Learning), an introductory DL course at Tech Mahindra to 50+ undergraduates from diverse academic backgrounds.
  • Received the OUTR Merit Scholarship (2020-21) for ranking 1st in my department for the last 2 UG years.
  • Beyond my work, I enjoy reading, sketching, cooking and playing team sports .
Template adopted from: 1 and 2