Use case in production
Stress testing Computer Vision models for robust deployment
Leonardo von Prellwitz
Global Enterprise Lead, Calypso AI
Leonardo is a graduate from the University of Chicago Booth School of Business (MBA) and Bocconi University (BSc.) and currently leads CalypsoAI's enterprise efforts for robust and trustworthy AI. Prior to joining CalypsoAI, Leo funded a fitness robotics start-up and helped Cisco's Emerging Technology group to establish their Smart Grid Business Unit.
Session description
Computer vision models are prone to underperformance or failure when exposed to edge-cases or adversarial attacks. We will provide an overview of tools and methodologies that can automate the stress-test workflow and simplify performance and risk assessment for business and technical decision-makers.