Fix tile bug & add more tests

[eugene123tw]

Summary

Good news, OV model maintains its accuracy without too much drop (1-2% on inst-seg). Interestingly, it even outperforms PyTorch model in terms of accuracy on detection models. Probably will look into it in another PR, but I think the risk for Geti is low at the moment.

OV Model	dataset	Image Size	Tile Size	Version	Torch Test/mAP50	OV Test/mAP_50	Inference Duration(s)
MaskRCNN-Eff	Vitens Aeromonas	3500 x 3500	1663 x 1663	2.x (PR)	0.9168	0.9061	59.016
MaskRCNN-R50	Vitens Aeromonas	3500 x 3500	1663 x 1663	2.x (PR)	0.8979	0.9052	72.322
MaskRCNN-Eff	Vitens Coliform	3500 x 3500	1663 x 1663	2.x (PR)	0.5311	0.5134	48.878
MaskRCNN-R50	Vitens Coliform	3500 x 3500	1663 x 1663	2.x (PR)	0.4854	0.5054	70.749
yolox_s	Vitens Aeromonas	3500 x 3500	1663 x 1663	2.x (PR)	0.6198	0.7948	37.497
yolox_s	Vitens Coliform	3500 x 3500	1663 x 1663	2.x (PR)	0.3721	0.42	23.833
yolox_tiny	Vitens Aeromonas	3500 x 3500	1663 x 1663	2.x (PR)	0.9354	0.9529	34.481
yolox_tiny	Vitens Coliform	3500 x 3500	1663 x 1663	2.x (PR)	0.5171	0.5943	21.262
yolox_l	Vitens Aeromonas	3500 x 3500	1663 x 1663	2.x (PR)	0.7258	0.7989	71.366
yolox_l	Vitens Coliform	3500 x 3500	1663 x 1663	2.x (PR)	0.43689	0.4206	51.628

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) 2023 Intel Corporation
# SPDX-License-Identifier: Apache-2.0