Case Study: Aerial Imagery Company achieves high-quality fast image annotation with Tasq.ai

How fast does it take to label an image with 3,000 objects

Aerial Imagery Company worked with Tasq.ai to tackle a difficult aerial image labeling challenge: annotating images containing up to 3,000 objects each, with accurate bounding boxes and object classifications. Their internal team found the process slow, tedious, and exhausting, and outsourcing through BPOs proved expensive and lower quality.

Tasq.ai solved the problem using its microtasking workflow platform, breaking each image into hundreds of small tasks, adding human validation, and using ML-assisted bounding box creation and classification. With more than 500 Tasgers contributing, the customer achieved high-quality annotations at much faster speed, reaching 99% accuracy and helping accelerate production readiness.