How to choose a machine learning model for the available data?

Question

Neural networks

The task is to detect one class of an object. As training data, there are 15000 256x256 images and the coordinates of the four corners of the bounding boxes that circle the object (checkerboard). The frames are not parallel to the image and are not rectangular. What is the best machine learning model to use for this task?

Answer 1

Are you going to write a recognizer from scratch yourself? Very unlikely. This means that you do not need to select a model, but look at how and with what tools people have been solving such a problem for a long time.
Here, for starters:
https://towardsdatascience.com/object-detection-te...
https://towardsdatascience.com/object-detection-al...
https://www.r-bloggers.com/2021 /09/object-detectio...
https://www.kdnuggets.com/2019/08/2019-guide-objec...
https://habr.com/ru/post/552298/
https://towardsdatascience .com/object-detection-wi...
https://towardsdatascience.com/how-to-detect-objec...
https://habr.com/en/company/jetinfosystems/blog/498652/
https:/ /towardsdatascience.com/real-time-object-de...
https://towardsdatascience.com/object-detection-si...
Well, then follow the links.