Evaluating explainable artificial intelligence methods for multi-label deep learning classification tasks in remote sensing
This work addresses the need for transparency in black-box deep learning models for remote sensing applications, providing insights into model behavior and dataset biases, but it is incremental as it applies existing XAI methods to new tasks.
The study evaluated ten explainable AI (XAI) methods for multi-label deep learning classification in remote sensing using BigEarthNet and SEN12MS datasets, finding that Occlusion, Grad-CAM, and Lime were the most interpretable and reliable, though they had limitations such as low resolution and high computational cost.
Although deep neural networks hold the state-of-the-art in several remote sensing tasks, their black-box operation hinders the understanding of their decisions, concealing any bias and other shortcomings in datasets and model performance. To this end, we have applied explainable artificial intelligence (XAI) methods in remote sensing multi-label classification tasks towards producing human-interpretable explanations and improve transparency. In particular, we utilized and trained deep learning models with state-of-the-art performance in the benchmark BigEarthNet and SEN12MS datasets. Ten XAI methods were employed towards understanding and interpreting models' predictions, along with quantitative metrics to assess and compare their performance. Numerous experiments were performed to assess the overall performance of XAI methods for straightforward prediction cases, competing multiple labels, as well as misclassification cases. According to our findings, Occlusion, Grad-CAM and Lime were the most interpretable and reliable XAI methods. However, none delivers high-resolution outputs, while apart from Grad-CAM, both Lime and Occlusion are computationally expensive. We also highlight different aspects of XAI performance and elaborate with insights on black-box decisions in order to improve transparency, understand their behavior and reveal, as well, datasets' particularities.