Case Study: xarvio achieves scalable, accurate crop identification with Labelbox

Using Labelbox to optimize crop production

Xarvio, part of BASF Digital Farming, develops the xarvio SCOUTING app used by farmers in 100+ countries (more than 3.2 million users) to identify weeds, pests and diseases via smartphone photos. To teach its computer-vision models to recognize thousands of crop issues across diverse regions, Xarvio faced the challenge of producing millions of high-quality labeled images and coordinating regional experts for accurate, scalable annotation—so the company adopted Labelbox’s training data/annotation platform to manage that process.

Using Labelbox, Xarvio uploads images into a central workflow where automated pre-categorization, rapid annotator tools and agriculture-expert pre-classification and quality-control checks produce vetted training data at scale. Labelbox accelerated iteration of training datasets, improved model accuracy and enabled new features in the SCOUTING app—contributions Xarvio credits with making “huge strides forward” and supporting its global user base and expanding radar-driven insights.

Xarvio

Marek Schikora

Global Product Owner