Towards Arbitrary-scale Histopathology Image Super-resolution: An Efficient Dual-branch Framework based on Implicit Self-texture Enhancement
This enables more flexible and accurate image analysis for medical professionals, though it is incremental as it adapts existing implicit neural network methods to a specific domain.
The paper tackles the problem of arbitrary-scale super-resolution for histopathology images, which previously only worked at fixed integer magnifications, by proposing a dual-branch framework with self-texture enhancement, and it outperforms existing methods on two public datasets.
Existing super-resolution models for pathology images can only work in fixed integer magnifications and have limited performance. Though implicit neural network-based methods have shown promising results in arbitrary-scale super-resolution of natural images, it is not effective to directly apply them in pathology images, because pathology images have special fine-grained image textures different from natural images. To address this challenge, we propose a dual-branch framework with an efficient self-texture enhancement mechanism for arbitrary-scale super-resolution of pathology images. Extensive experiments on two public datasets show that our method outperforms both existing fixed-scale and arbitrary-scale algorithms. To the best of our knowledge, this is the first work to achieve arbitrary-scale super-resolution in the field of pathology images. Codes will be available.