We use top-1 error rate as the evaluation metric. For each image, an algorithm will produce 1 label. For this competition each image has one ground truth label. For a given image, if the ground truth label is same as the predicted label, then the error for that image is 0, otherwise it is 1. The final score is the error averaged across all images. See
[here](https://github.com/visipedia/fgvcx_flower_comp#evaluation)
for additional information.
Submission Format
For each image in the test set, you must predict 1 class label. The csv file should contain a header and have the following format:
id,predicted
12345, 23
67890, 42
The
id
column corresponds to the test image id. The
predicted
column corresponds to the predicted category id. You should have one row for each test image.