Title: Deployment‑First Approach to Agricultural Vision Intelligence
Abstract
Agriculture is undergoing a digital transformation driven by advances in computer vision, remote sensing, and artificial intelligence. These technologies now operate across unprecedented scales—from cellular microscopy and laboratory imaging to drones and satellite‑based Earth observation. However, achieving real‑world impact requires more than accurate models; it demands systems designed for deployment from day one. This talk presents a deployment‑first approach to agricultural vision intelligence, highlighting how end‑to‑end thinking enables scalable, usable, and impactful AI solutions. Drawing on real examples across laboratory phenotyping, proximal field sensing, airborne imaging, and space‑borne analytics, the talk explores how modern vision systems are built under constraints such as limited annotations, massive data volumes, heterogeneous sensors, and edge‑compute requirements. Key topics include self‑supervised learning, few‑shot adaptation, multimodal vision models, and platform‑aware deployment strategies. Attendees will gain insight into how AI can transition from research prototypes to production systems that deliver measurable value to farmers, agronomists, and agricultural researchers.
Speaker Bio
Sudhir Sornapudi is a Principal Investigator and leads in Advanced Vision Intelligence team at Corteva, with over a decade of experience applying computer vision and deep learning to real‑world problems. He holds a Ph.D. in Computer Engineering from Missouri University of Science and Technology and has published at CVPR and related venues. His work spans digital pathology, digital agriculture, remote sensing, laboratory imaging, and large‑scale AI deployment, with a focus on translating research into production‑ready systems.
