Continual image super-resolution (CISR) aims to efficiently adapt a pre-trained model to a variety of tasks while retaining knowledge from previously learned tasks, minimizing the need for intensive independent training. The primary challenges include catastrophic forgetting due to varying data distributions and degradation types, along with the necessity for high adaptability. While prompt-based continual learning has proven effective in image classification, its direct application to super-resolution (SR) often fails to meet the demands for detailed pixel-level restoration and domain discrimination in low-level characteristics. To address these challenges, we propose Learning Prompt Adapters (LPA), which dynamically generates pixel-wise prompts through a combination of multi-granularity prompt bases and identities. By adaptively integrating these prompts into the Transformer architecture, we effectively improve the model's performance on fine-grained details in super-resolution tasks, as well as enhancing the model's adaptability to new tasks and preserving knowledge from previous ones. Through organizing the low-rank prompt bases with specific identities, we set up an effective solution to managing cross-task differences and enhancing prompt richness. Extensive experiments on benchmarks comprising the NYU, RealSR, DIV2K, REDS, and MANGA109 datasets with diverse degradation types demonstrate that LPA significantly outperforms existing continual learning methods. Codes of this paper are available at: https://github.com/dummerchen/LPA.

Realistic image super-resolution (SR) focuses on transforming real-world low-resolution (LR) images into high-resolution (HR) ones, handling more complex degradation patterns than synthetic SR tasks. This is critical for applications like surveillance, medical imaging, and consumer electronics. However, current methods struggle with limited real-world LR-HR data, impacting the learning of basic image features. Pre-trained SR models from large-scale synthetic datasets offer valuable prior knowledge, which can improve generalization, speed up training, and reduce the need for extensive real-world data in realistic SR tasks. In this paper, we introduce a novel approach,Dual-domain Adaptation Networks, which is able to efficiently adapt pre-trained image SR models from simulated to real-world datasets. To achieve this target, we first set up a spatial-domain adaptation strategy through selectively updating parameters of pre-trained models and employing the low-rank adaptation technique to adjust frozen parameters. Recognizing that image super-resolution involves recovering high-frequency components, we further integrate a frequency domain adaptation branch into the adapted model, which combines the spectral data of the input and the spatial-domain backbone's intermediate features to infer HR frequency maps, enhancing the SR result. Experimental evaluations on public realistic image SR benchmarks, including RealSR, D2CRealSR, and DRealSR, demonstrate the superiority of our proposed method over existing state-of-the-art models.

Sea surface temperature (SST) data derived from infrared satellites are frequently missing due to cloud cover. To address this challenge, we propose a novel interpolation model, the Data Interpolation Network with a Collaborative Dual Encoder (DINCoDE). The model adopts a hybrid dual-encoder architecture that integrates a Convolutional Neural Network (CNN) and a Transformer encoder, enabling the joint extraction of local and global spatiotemporal features. A composite loss function is employed to ensure both pixel-level accuracy and structural fidelity in the reconstructed data. Experimental results on the OSTIA SST dataset (N=6,575) demonstrate that DINCoDE outperforms the DINCAE and DINEOF models across all evaluation metrics (RMSE = 0.25°C, MAE = 0.16°C, R² = 0.99, SSIM = 0.99) and generates more physically realistic spatial gradient fields. Notably, the pre-trained model demonstrated robust performance when applied to the Advanced Himawari Imager (AHI) SST dataset (RMSE = 0.38°C, MAE = 0.27°C, R² = 0.99), highlighting its strong generalization ability and potential for operational deployment. Moreover, our analysis reveals that the intrinsic spatial complexity and variability of the SST field are key factors influencing reconstruction performance. DINCoDE offers a high-accuracy, robust, and generalizable method for reconstructing SST data gaps caused by cloud occlusion, providing a reliable solution for SST data reconstruction tasks in oceanographic applications.