* remove get_image and change it to get_images * add config attribute for color_source_name * refactor: split image retrieval to avoid unnecessary tensor creation
* remove get_image and change it to get_images * add config attribute for color_source_name * refactor: split image retrieval to avoid unnecessary tensor creation