Q: What do I need to use this API?

You can train a specialized model only using a short video from your environment.

Q: What computer vision tasks are supported?

Right now we are supporting object detection for all COCO object categories. We are working on Adaptive pose estimation for our next release.

Q: What inference devices are supported?

You can run our optimized model on NVIDIA GPUs and Jetson devices, Edge TPU, and CPU devices.

Q: How much does it cost?

Our service is currently in alpha and is free of charge. More details on the pricing will be available later.

Q: What happens to my data?

Your input video data is used to create a model specifically for you. We don't share your videos or use your videos for any other purposes.