Zebrafish Tracking

Zebrafish tracking uses a pose detection neural network to infer 8 key points on each fish and using the resulting cartesian coordinates determine various kinematic parameters such as distance traveled, speed, tail bend angle, and swim heading.

The zebrafish tracking workflow consists of two parts: acquisition and analysis. During acquisition users can flexibly configure recording duration, frame rate, and integrated stimuli (flash, vibration, and tap). The analysis portion automatically detects the well plate in use (24, 48, 96 well plate) and users may select which output metrics they would like to receive.