The everything generation of official Demo

I use SAM to generate every objects in image based on SAM source code. But I found the results of SAM offical demo are better than those of SamAutomaticMaskGenerator of souce code. Could you tell me Why? Did you process specially the selection points in your official demo?

Thanks for your answers.

Compare the official demo entry point with SamAutomaticMaskGenerator in the SAM source code, focusing first on how selection points are handled. Determine whether the two paths use different inputs or processing, and document the cause of the differing results as the completion criterion.

Written by the indexing model from the issue text.