 This paper proposes a novel semi-supervised hyperspectral image classification method based on generative adversarial networks, GANs. It uses a 3D bilateral filter, 3DBF, to extract spectral spatial features from the hyperspectral images, then trains a GAN on these features to generate additional data points for the classification model. This allows the model to learn from both labeled and unlabeled data, making it more robust and efficient than traditional supervised models. Experiments on three benchmark hyperspectral datasets demonstrate the effectiveness of the proposed method, especially when there are few labeled samples available. This article was authored by Jihee, Hanlu, Yiwen Wang, and others.