In this session, I will share best practices and insights into vision-guided robotics, highlighting an innovative solution that integrates few-shot learning with NVIDIA FoundationPose and NVIDIA cuMotion. Our approach simplifies workflows by reducing the need for extensive annotations and lengthy model training, which are common in traditional AI-based vision-guided robotics systems. The implementation is made even more accessible through an intuitive, no-code drag-and-drop web interface.

Discover our few-shot learning architectures with NVIDIA FoundationPose and cuMotion specifically designed for vision-guided robotics.
Gain insights from our real-world experiences in material handling applications.
Understand the cost-benefit analysis behind our implemented use cases.