Hello, I'm pleased that you've made the source code open for this article. I am particularly interested in the ColoredCatsDogs dataset, but I have identified a couple of issues. Firstly, in the code, it is stated that all train, val, and test images should be placed in the 'ccd' directory under the 'images' folder. However, this requirement is not mentioned in the README. Secondly, I noticed discrepancies in the split JSON file for the ColoredCatsDogs dataset provided on Baidu Netdisk. It appears that there might be errors in the file, or the dataset might be incomplete, as some images are not present in the dataset.
These are my concerns, and I would appreciate it if you could address them. Thank you!
Hello, I'm pleased that you've made the source code open for this article. I am particularly interested in the ColoredCatsDogs dataset, but I have identified a couple of issues. Firstly, in the code, it is stated that all train, val, and test images should be placed in the 'ccd' directory under the 'images' folder. However, this requirement is not mentioned in the README. Secondly, I noticed discrepancies in the split JSON file for the ColoredCatsDogs dataset provided on Baidu Netdisk. It appears that there might be errors in the file, or the dataset might be incomplete, as some images are not present in the dataset.
These are my concerns, and I would appreciate it if you could address them. Thank you!