multi-object retrieval

As far as I know, most of the current image retrieval is for a single object or several objects. But for the field of autonomous driving, a picture taken is multi-object, for example, a picture contains people, cars, fences, buildings, trees, etc. Only using [cls_token] will lose a lot of object information. Are there any suggestions for multi-object retrieval?

The issue names no files, tests, or entry points. Start by locating the existing image-retrieval path and how [cls_token] is used, then determine whether the proposed multi-object case has an agreed implementation direction; done would require a concrete, validated approach rather than general suggestions.

Written by the indexing model from the issue text.