Would it be significant? I think most computer vision feature descriptors can be run over multiple images in parallel. It'd be neat to reduce the entropy of po

Would it be significant? I think most computer vision feature descriptors can be run over multiple images in parallel. It'd be neat to reduce the entropy of possible photos by using EXIF data (location, etc.)