File size: 451 Bytes
bfd2e55 | 1 2 3 4 5 6 7 8 | # Oxford Pets Models
Contains the two models for Oxford Pets.
One is a self-trained image classifier, the other is a self-trained image masker.
The ensemble combination of these models is a simple single-object semantic segmentation model.
They had around 67% accuracy, and 58% accuracy respectively.
Because they were not fine-tuned and trained for such few epochs, these are actually quite solid results.
Further training will update these results. |