This notebook demonstrates how the use attention center model with OpenVINO. This model is in the TensorFlow Lite format, which is supported in OpenVINO now by TFLite frontend. Check out this article to find more information about this model. The attention-center model takes an RGB image as input and return a 2D point as result, which is the predicted center of human attention on the image.
The tutorial consists of the following steps:
- Downloading the model
- Loading the model and make inference with OpenVINO API
- Run Live Attention Center Detection
If you have not installed all required dependencies, follow the Installation Guide.