3 ms·

Can someone explain the meaning of labelled vs unlabelled in this context? What kind of information would the labels carry? Did they have depth maps for all 62

• yorwba 3y ago
  They explain in the paper that they used 1.5 million images with known depth maps (labels) to train a teacher model, and then used the teacher model to create pseudolabels (inferred depth maps) for the full dataset. Then they trained a student model to recover those pseudolabels from distorted versions of the original images.
  • throwaway81523 3y ago
    Was that better than running the teacher model on the distorted images directly?