Foundation-model post-training
Supervised fine-tuning, preference optimization, and reinforcement learning for multimodal models.
Multimodal foundation models for video
Divyanshu Mishra
Research Scientist II Amazon Science
Visiting Researcher University of Oxford
My current work at Amazon Science focuses on post-training foundation models and building reliable reasoning systems. I am particularly interested in supervised fine-tuning, reinforcement learning, and multimodal models that can understand video.
I completed my DPhil in Engineering Science at the University of Oxford, supervised by Professor Alison Noble and supported by the Athena-Bronze Scholarship. My doctoral research centered on long-video understanding, including temporal localization, self-supervised representation learning, and multimodal learning. I developed these methods primarily for fetal ultrasound, with the goal of supporting earlier and more reliable detection of congenital heart disease.
Before joining Amazon full-time, I interned at Amazon Science as an Applied Scientist II. There, I developed text-only training strategies for Video-LLMs to reduce their dependence on large paired video-text datasets.
Before my DPhil, I was a Data Scientist at THSTI, Government of India, where I worked under Professor Shinjini Bhatnagar. Together with clinicians and public health researchers, I developed models for preterm birth prediction, gestational-age estimation, and privacy-preserving ultrasound.
Across these roles, the common thread in my work is learning effectively from limited supervision and translating research into systems that remain useful in real-world settings. I am always happy to connect about research, collaborations, and new ideas.
Research
Current focus
Long-video understanding
Temporal grounding, representation learning, and world models for long, unstructured video.
AI for healthcare
Reliable multimodal systems for clinical imaging, early detection, and decision support.
Experience
7+ years in AI research
Seven years spanning applied research in industry and doctoral work at Oxford.
Research Scientist II
Amazon Science · Seattle
Visiting Researcher
University of Oxford
DPhil (PhD), Engineering Science
University of Oxford · Noble Lab
Applied Scientist II Intern
Amazon Science · Berlin
Data Scientist, Computer Vision
THSTI · Government of India
Selected work
Research highlights
Methods that learn useful temporal structure from complex, real-world video.
NeurIPS 2025
DISCOVR
Self-supervised video representations through online cluster distillation.
MICCAI 2025
STUD + DiVMerge
Zero-shot anomaly detection with privacy-preserving model merging.
Medical Image Analysis 2025
TIER-LOC
Visual query-based localization in long fetal ultrasound video.
AAAI 2025
MCAT
Class-aware token transformers for anatomical clip localization.
What’s new
Recent updates
Research output
First- & co-first-author papers
medRxiv 2026
Development and External Validation of Deep Learning Models for Spontaneous Preterm Birth Prediction from Mid-Trimester Cervical Ultrasound
Radhika Chanian*, Divyanshu Mishra*, Rahul Jain, Nikhil Sharma, Ashok Khurana, Reva Tripathi, Abhinav Jain, GARBH-Ini study group, Nitya Wadhwa, J. Alison Noble, Ramachandran Thiruvengadam, Bapu Koundinya Desiraju, and Shinjini Bhatnagar
Manuscript under review
Grounding in the Dark: Investigating Text-Only Video-LLM Training Strategies for Zero-Shot Video Temporal Localization
Divyanshu Mishra, S. Sternig, R. Shetty, and Erhan Gundogdu
MICCAI 2025 · Co-first author
Self-supervised Normality Learning and Divergence Vector-guided Model Merging for Zero-shot Congenital Heart Disease Detection in Fetal Ultrasound Videos
Pramit Saha*, Divyanshu Mishra*, Netzahualcoyotl Hernandez-Cruz, Olga Patey, Aris T. Papageorghiou, Yuki M. Asano, and J. Alison Noble
Medical Image Analysis 2025
TIER-LOC: Visual Query-based Video Clip Localization in Fetal Ultrasound Videos with a Multi-Tier Transformer
Divyanshu Mishra, Pramit Saha, He Zhao, Netzahualcoyotl Hernandez-Cruz, Olga Patey, Aris T. Papageorghiou, and J. Alison Noble
Medical Image Analysis 2025 · Co-first author
HarmonicEchoNet: Leveraging Harmonic Convolutions for Automated Standard Plane Detection in Fetal Heart Ultrasound Videos
Md Mostafa Kamal Sarker*, Divyanshu Mishra*, Mohammad Alsharid*, Netzahualcoyotl Hernandez-Cruz, Rahul Ahuja, Olga Patey, Aris T. Papageorghiou, and J. Alison Noble
AAAI 2025
MCAT: Visual Query-Based Localization of Standard Anatomical Clips in Fetal Ultrasound Videos using Multi-Tier Class-Aware Token Transformer
Divyanshu Mishra, Pramit Saha, He Zhao, Netzahualcoyotl Hernandez-Cruz, Olga Patey, Aris Papageorghiou, and J. Alison Noble
medRxiv 2024
Development and External Validation of an Ultrasound Image-Based Deep Learning Model to Estimate Gestational Age in the Second and Third Trimesters of Pregnancy Using Data from the GARBH-Ini Cohort: A Prospective Cohort Study in North Indian Population
Divyanshu Mishra, Varun Chandramohan, Nikhil Sharma, Mudita Gosain, Nitya Wadhwa, Uma Chandra Mouli Natchu, GARBH-Ini study group, Ashok Khurana, J. Alison Noble, Ramachandran Thiruvengadam, Bapu Koundinya Desiraju, and Shinjini Bhatnagar
MICCAI 2024
STAN-LOC: Visual Query-Based Video Clip Localization for Fetal Ultrasound Sweep Videos
Divyanshu Mishra, Pramit Saha, He Zhao, Olga Patey, Aris T. Papageorghiou, and J. Alison Noble
MICCAI 2023
Dual Conditioned Diffusion Models for Out-of-Distribution Detection: Application to Fetal Ultrasound Videos
Divyanshu Mishra, He Zhao, Pramit Saha, Aris T. Papageorghiou, and J. Alison Noble