Write a comment

This manuscript presents an exploration of bias detection in deep learning, applying Grad-CAM analysis in plant species classification. Among its strengths is the topic, focuses on Explainable AI (XAI) and bias detection, which is highly relevant in domains such as agriculture. The manuscript evaluates five CNN architecture, offering a broad comparative view and the dataset, parameters and metrics are sufficiently described. Also, the manuscript is well-written and constitutes an original contribution to the literature with practical implications.

On the other hand, further explanation is needed in the introduction about CNNs and Grad-CAM, to the unfamiliar readers of the manuscript. In the materials and methods section, there is no provided information about the resources, both hardware and software, to perform the tasks, which are essential for reproducibility purposes. Finally, graphs and figures should be enlarged to be more distinct, units of measurements and error bars and citations of figures in the text should be added and in the Table 1, a legend would be helpful about the abbreviations used.

Competing interests

The author declares that they have no competing interests.

Use of Artificial Intelligence (AI)

The author declares that they used generative AI to come up with new ideas for their review.