🤖 AI Summary
This work addresses the issue of “condition collapse” in digital histopathology image generation, which arises when using pretrained vision foundation models as conditioning signals. The authors propose a Riemannian flow matching–based generative framework that models normalized patch-token features as latent variables on the unit hypersphere. Their approach introduces, for the first time, a spherical-aware bridging stochastic perturbation mechanism and an anisotropic decoder, integrated within a Diffusion Transformer (DiT) architecture. Generation is further guided by a decoding strategy based on the directional energy of the Jacobian matrix of the velocity field. Evaluated on breast and colorectal cancer datasets, the method achieves state-of-the-art performance in both reconstruction fidelity and generation diversity, significantly outperforming existing approaches.
📝 Abstract
Synthetic histopathology image generation addresses critical challenges in computational pathology, including patient privacy and the growing need for large-scale training data for foundation models. Latent diffusion models have dominated the image generation domain, with recent works emphasizing that the choice of latent space is critical to the quality of generated images. Existing state-of-the-art generative models in histopathology use pretrained Vision Foundation Models (VFMs) as conditioning signals, and we observe that this leads to "conditioning collapse," where the conditioning signal dominates the latent space and lowers the quality and diversity of generated samples. Therefore, we instead use pretrained histopathology VFMs as the latent space itself, leveraging their patch-token features that encode rich semantic information. We empirically show that these features are $\ell_2$-normalized and lie on the unit hypersphere $\mathcal{S}^{d-1}$ with strong angular dominance and intrinsic curvature, making them naturally suited for a Riemannian formulation. We therefore present STREAM, the first framework to apply Riemannian flow matching in the pathology domain. STREAM consists of two stages: 1) a bridge-type stochastic perturbation that establishes per-token rectifiability on $\mathcal{S}^{d-1}$ for training a Diffusion Transformer (DiT) in latent space, and 2) a novel anisotropic decoder that allocates robustness to low-energy directions of the velocity-field Jacobian while preserving fidelity along its high-energy directions. Together, STREAM achieves state-of-the-art reconstruction and generation performance on breast and colorectal cancer datasets. The code will be publicly released upon acceptance.