Dataset and Results

[songeza]

1.The original RUWI training set and test/validation set consisted of 400 and 300 images, respectively. Could you please explain why in this paper the training and test sets consist of 525 and 175 images, respectively? Could you please provide more details about the dataset, such as the specific format of the train.txt and val.txt files? 2.To enhance the impact of the paper and avoid reinventing the wheel, it has been customary for your team (led by Prof. Huchuan Lu) to share results of comparative methods in previous works. Can you confirm if the authors have plans to do the same in this paper?