The effectiveness of spectral-spatial feature learning is crucial for the hyperspectral image (HSI) classification task. Diffusion models, as a new class of groundbreaking generative models, have the ...

Diffusion models can enhance features and aggregate similar ones, but they have a high computational cost due to the sampling process. We combine contrastive learning and diffusion models by training ...

Counterfactual explanations (CEs) aim to enhance the interpretability of machine learning models by illustrating how alterations in input features would affect the resulting predictions. Common CE ...

The 336 FOV yielded a classification accuracy of 91.15%, surpassing the real dataset’s 80.06% and the combined dataset’s 88.21%. In computational pathology, generative models can enhance data sharing ...

The essence of the DDPM diffusion model is learning a “denoising ... In cases of limited samples, transfer learning is applied to transfer the model’s generic features from other pre-trained networks ...

The study comprehensively evaluates diffusion-based representation learning across various datasets and tasks, shedding light on their potential derived solely from images. The proposed model ...

In its traditional model, Stable Diffusion 1.4 can generate artwork thanks to its acquired ability to make relevant statistic associations between images and related words.

If you find the code helpful for your research, please cite: J. Zhou et al., "Exploring Multi-Timestep Multi-Stage Diffusion Features for Hyperspectral Image Classification," in IEEE Transactions on ...