A good way to see the basics of computer vision. This gives all the building blocks needed.

A look back at the limitations of deep learning in the context of computer vision. We're better at avoiding over fitting nowadays but the shallowness of the available data is still a problem.

So much data trapped in PDFs indeed... Unfortunately VLM are still not reliable enough to be unleashed without tight validation of the output.

Those brand new models keep failing at surprisingly simple tasks.

Looks like an interesting pipeline for multi-view stereo reconstruction.