Oxford IIIT Pet Dataset (fixed)

Dataset description

The Oxford-IIIT Pet Dataset is a 37 category pet dataset with roughly 200 images for each class created by the Visual Geometry Group at Oxford. The images have a large variations in scale, pose and lighting. All images have an associated ground truth annotation of breed, head ROI, and pixel level trimap segmentation. [kaggle.com]

Original dataset was downloaded from https://www.robots.ox.ac.uk/~vgg/data/pets/.

Fixed images

The original dataset contains a few damaged images. We fixed them and uploaded into this repository. These images are mentioned below, as well as the steps of fixing them.

1. Abyssinian_34.jpg

    File header: GIF image data, version 89a, 250 x 202
   

   File was converted to jpeg image format.

2. Egyptian_Mau_139.jpg

    File header: GIF image data, version 89a, 350 x 250
   

   File was converted to jpeg image format.

3. Egyptian_Mau_145.jpg

    File header: GIF image data, version 89a, 216 x 188
   

   File was converted to jpeg image format.

4. Egyptian_Mau_167.jpg

    File header: GIF image data, version 89a, 183 x 27
   

   File was converted to jpeg image format.

5. Egyptian_Mau_177.jpg

    File header: GIF image data, version 87a, 300 x 214
   

   File was converted to jpeg image format.

6. beagle_116.jpg

    Corrupt JPEG data: premature end of data segment
   

   File was loaded using OpenCV and saved again as jpeg image. It is not possible to fix damaged regions easily.

7. chihuahua_121.jpg

    Corrupt JPEG data: 240 extraneous bytes before marker 0xd9
   

   File was loaded using OpenCV and saved again as jpeg image. It is an easy way how to clean file from purposeless bytes.