Online
Hosted by Amsterdam AI Machine Learning and Computer Vision Meetup
Explore how computer vision, edge AI and drones are tackling practical challenges in agriculture, wildfire response and environmental field sampling. This online meetup takes place on 5 November 2026, 9:00–11:00 AM PST (17:00–19:00 UTC). Hosted by the Amsterdam AI Machine Learning and Computer Vision Meetup, it is aimed at AI/ML researchers, computer vision practitioners and people working across agriculture, robotics and environmental technology. The online access link is visible to attendees. Talks - Micro-scale soybean seed damage segmentation, Saurav Upadhyaya will present a two-stage U-Net framework with a ResNet34 encoder for identifying wrinkles, dark spots and surface damage in high-resolution seed imagery. The research uses 77,000 individual seed images and specialist models reporting validation accuracies of 94% and 98.53%. - Wildfire detection at the edge, Maxime Carriere, co-founder of Kernwerk, will trace the journey from a compact camera-based smoke detection model to terrain-, weather- and history-aware alert scoring, followed by drone confirmation. The talk will examine model compression, edge deployment and trust in automated detections. - Can a drone sample as well as a human?, Rohan Talwadia, co-founder and CTO of LabGPT, will share lessons from 5,000 drone flights supporting soil, water and vegetation sampling. Topics include multispectral and thermal sensing, route planning, GPS degradation, chain of custody, regulatory comparability and calibrated uncertainty. Expect a technical programme grounded in deployed systems, measurable results and real-world operational constraints rather than high-level AI theory. Register for the Visual AI Agriculture Meetup
