[_wp_]

> To use people's random images for training, they would have to be manually annotated by a human (e.g. facial boxes, eyes, nose, mouth, ears drawn in).

That's not true. There is a large and growing body of research on semi-supervised, self-supervised, and unsupervised learning that can take advantage of these unlabelled images.

[arghwhat]

Different learning techniques have different applications. I do not believe those techniques are applicable to the hypothetical use-cases of this dataset.

Perhaps semi-supervised could be utilized, which reduces the required annotation by some factor k, but still leaves it as a function of the dataset.

Self-supervised basically replaces human annotation with machine annotation, making it only applicable to a small subset of tasks in which this is possible (e.g. you could train "guess time from picture" using EXIF timestamp).

Unsupervised is only applicable to very specific tasks.