Introduction to Smart Capture

📘

Welcome to the documentation for Scale Smart Capture!

Explore workflows and tutorials using the side bar or search!

Edge devices such as vehicles, drones or robots are resource constrained and don’t have sufficient storage or network bandwidth to store or stream all captured data. As a result, valuable edge case data is lost making it hard to improve their model performance.

To address this, Smart Capture enables ML teams to intelligently capture the most valuable data from their edge devices to improve their models.

Smart Capture has two components: First, an on-device SDK to evaluate the triggers and decide what data should be retained. And second, a web interface and API for managing devices, and triggers which define under what conditions data should be captured.