The notebook will cover the following topics:
- Explanation of a saliency map and how it can be used.
- Overview of the CLIP neural network and its usage in generating saliency maps.
- How to split a neural network into parts for separate inference.
- How to speed up inference with OpenVINO™ and asynchronous execution.
A saliency map is a visualization technique that highlights regions of interest in an image. For example, it can be used to explain image classification predictions for a particular label. Here is an example of a saliency map that we will get in this notebook:
This is a self-contained example that relies solely on its own code.
We recommend running the notebook in a virtual environment. You only need a Jupyter server to start.
For details, please refer to Installation Guide.