Drug repositioning can effectively reduce research and development costs and accelerate time to market by identifying new indications for existing drugs. In recent years, deep learning-based methods have achieved remarkable progress in the field of drug repositioning. However, current approaches still suffer from several limitations. First, many methods simply concatenate structural information and association information without fully capturing their intrinsic relationships, leading to suboptimal information fusion. Second, most models lack mechanisms for extracting high-order structural information and aligning heterogeneous and homogeneous features, which limits the model's expressiveness and predictive performance. To address these limitations, we propose TransFusionDR, a framework for drug repositioning via contrastive and high-order feature fusion with transformers. We employ a Graph Transformer to extract deep structural features from drug-drug and disease-disease graphs, and utilize a Heterogeneous Graph Transformer to capture semantic information from the drug-disease association graph. These features are then refined through contrastive learning to enhance semantic consistency and improve information fusion quality. We further introduce a Transformer Encoder to deeply integrate the homogeneous and heterogeneous features by dynamically modeling semantic dependencies, enabling the extraction of high-order interactions and achieving more effective feature alignment. Experimental results on two benchmark datasets demonstrate that the proposed framework significantly improves prediction accuracy and robustness in drug repositioning tasks, outperforming state-of-the-art methods.

Existing research in low-level vision has shifted its focus from "one-by-one" task-specific methods to "all-in-one" multi-task unified architectures. However, current all-in-one image restoration approaches primarily aim to improve overall performance across a limited number of tasks. In contrast, how to incrementally add new image restoration capabilities on top of an existing model — that is, task-incremental learning — has been largely unexplored. To fill this research gap, we propose a minimalistic and universal paradigm for task-incremental learning called MINI. It addresses the problem of parameter interference across different tasks through a simple yet effective mechanism, enabling nearly forgetting-free task-incremental learning. Specifically, we design a special meta-convolution called MINI-Conv, which generates parameters solely through lightweight embeddings instead of complex convolutional networks or MLPs. This not only significantly reduces the number of parameters and computational overhead but also achieves complete parameter isolation across different tasks. Moreover, MINI-Conv can be seamlessly integrated as a plug-and-play replacement for any convolutional layer within existing backbone networks, endowing them with incremental learning capabilities. Therefore, our method is highly generalizable. Finally, we demonstrate that our method achieves state-of-the-art performance compared to existing incremental learning approaches across five common image restoration tasks. Moreover, the near forgetting-free nature of our method makes it highly competitive even against all-in-one image restoration methods trained in a full-supervised manner. Our code is available at https://github.com.

This paper proposes a fast and general-purpose image restoration method. The key idea is to achieve few-step or even one-step inference by conducting consistency distilling or training on a specific mean-reverting stochastic differential equations. Furthermore, based on this, we propose a novel linear-nonlinear decoupling training strategy, significantly enhancing training effectiveness and surpassing consistency distillation on inference performance. This allows our method to be independent of any pre-trained checkpoint, enabling it to serve as an effective standalone image-to-image transformation model. Finally, to avoid trivial solutions and stabilize model training, we introduce a simple origin-guided loss. To validate the effectiveness of our proposed method, we conducted experiments on tasks including image deraining, denoising, deblurring, and low-light image enhancement. The experiments show that our method achieves highly competitive results with only one-step inference. And with just two-step inference, it can achieve state-of-the-art performance in low-light image enhancement. Furthermore, a number of ablation experiments demonstrate the effectiveness of the proposed training strategy. our code is available at https://github.com/XiaoxuanGong/IR-CM.