Fix spatial resolution for saliency_maps in torchvision_models

[GalyaZalesskaya]

Summary

• Fixed spatial resolution for saliency_maps in torchvision_models. CVS-137552
• get_feature_vector -> feature_vector_fn renaming for consistency. Reguest

Currently, in torchvision models the features are passed to the explainer after passing AvgPool2D layer, that causes the final saliency map to lose spatial resolution (e.g. (2, 3, 1, 1) instead of (2, 3, 7, 7)).
This PR breaks backbone into feature_extractor and avgpool and passes features after feature_extractor to explainer.
Also I added tests to check spatial resolution of the resulting saliency maps.

Issue: CVS-137552

How to test

Checklist

• I have added unit tests to cover my changes.​
• I have added integration tests to cover my changes.​
• I have added e2e tests for validation.
• I have added the description of my changes into CHANGELOG in my target branch (e.g., CHANGELOG in develop).​
• I have updated the documentation in my target branch accordingly (e.g., documentation in develop).
• I have linked related issues.

License

• I submit my code changes under the same Apache License that covers the project.
  Feel free to contact the maintainers if that's a concern.
• I have updated the license header for each file (see an example below).

# Copyright (C) 2024 Intel Corporation
# SPDX-License-Identifier: Apache-2.0

[GalyaZalesskaya]

@vinnamkim Here is the fix of the issue you reported