Yes, the idea works and was explored with dreambooth/textual inversion for image diffusion models.
https://dreambooth.github.io/ https://textual-inversion.github.io/
https://dreambooth.github.io/ https://textual-inversion.github.io/
InstantID (https://replicate.com/zsxkib/instant-id) fixes that issue.
If you just want a face, InstandID/Pulid work - but it’s not going to be very varied. Doing actual training means you can get any perspective, lighting, style, expression, etc - and have the whole body be accurate.