Label-free Anomaly Detection in Aerial Agricultural Images with Masked Image Modeling
This work addresses the challenge of early anomaly detection in agricultural UAV images for farmers, offering a label-free method that generalizes across multiple stress types, though it is incremental as it builds on existing self-supervised techniques.
The paper tackled the problem of detecting various stresses in agricultural fields from aerial images by framing it as an anomaly detection task, using a self-supervised masked image modeling approach with an anomaly suppression loss to eliminate the need for labeled normal data, resulting in improved mIOU scores on the Agriculture-Vision dataset compared to prior state-of-the-art methods.
Detecting various types of stresses (nutritional, water, nitrogen, etc.) in agricultural fields is critical for farmers to ensure maximum productivity. However, stresses show up in different shapes and sizes across different crop types and varieties. Hence, this is posed as an anomaly detection task in agricultural images. Accurate anomaly detection in agricultural UAV images is vital for early identification of field irregularities. Traditional supervised learning faces challenges in adapting to diverse anomalies, necessitating extensive annotated data. In this work, we overcome this limitation with self-supervised learning using a masked image modeling approach. Masked Autoencoders (MAE) extract meaningful normal features from unlabeled image samples which produces high reconstruction error for the abnormal pixels during reconstruction. To remove the need of using only ``normal" data while training, we use an anomaly suppression loss mechanism that effectively minimizes the reconstruction of anomalous pixels and allows the model to learn anomalous areas without explicitly separating ``normal" images for training. Evaluation on the Agriculture-Vision data challenge shows a mIOU score improvement in comparison to prior state of the art in unsupervised and self-supervised methods. A single model generalizes across all the anomaly categories in the Agri-Vision Challenge Dataset