3D Gaussian Splatting (3DGS) has attracted considerable attention due to its remarkable rendering efficiency and superior visual quality. However, typical implementations struggle with ultra-high-resolution imagery due to excessive GPU memory demands, a critical limitation particularly in domains such as aerial mapping and remote sensing, where fine-grained details are crucial for accurate 3D reconstruction. In this paper, we introduce InfiniGS, an enhanced Gaussian Splatting framework that scales up training image resolution while maintaining fidelity and efficiency, enabling reconstruction from ultra-high-resolution images without encountering out-of-memory (OOM) issues. Our framework consists of two modules; firstly, we propose a Multi-Resolution Training strategy that guides the optimization from recovering global structures toward fine-grained details, while mitigating erosion artifacts during resolution transitions. Secondly, we introduce an Image Block Partitioning and Asynchronous Optimization scheme, which significantly reduces GPU memory overhead. Most importantly, we reveal the scaling rule for the densification threshold of Gaussian primitives and the hyperparameters of the Adam optimizer during resolution transitions and image partitioning. This ensures consistently high rendering quality regardless of the multi-resolution factor or partitioning strategy employed. Extensive experiments demonstrate that InfiniGS achieves superior Novel View Synthesis (NVS) for high-resolution scenes, while requiring less memory and shorter training time. To the best of our knowledge, InfiniGS is the first method capable of reconstructing scenes with resolutions up to 10 K and beyond, at full resolution and high fidelity, without exceeding the memory capacity of mainstream powerful GPUs (e.g., NVIDIA RTX 3090 with 24 GB VRAM).

Clouds in optical remote sensing images (ORSI) significantly limit image utilization. Traditional cloud removal methods using single or multi-temporal data sources struggle to ensure reliable reconstruction for thick cloud areas. Synthetic Aperture Radar (SAR) images are increasingly used to recover information obscured by clouds, but their performance in cloud-obscured regions is unstable. Therefore, an adaptive cloud removal network for remote sensing images, named DG2-TCR, is proposed based on SAR-driven dual-flow fusion guidance (DFG). DG2-TCR uses SAR and ORSI to construct DFG, including local spatial-spectral feature reconstruction (LSSFR) flow and global texture feature compensation (GTFC). LSSFR, driven by ORSI and SAR, efficiently extracts useful features in non-cloud areas and focuses on local information reconstruction using the designed spatial-spectral features inference reconstruction block (SSIRB). Based on SAR images, GTFC guides the compensation of global texture information. DFG can adaptively extract features and reconstruct missing information from local and global scales. The public SEN12MS-CR-TS dataset is divided into four sub-datasets with different coverage to evaluate the recovering capability in varying clouds. Experiments show that the PSNR, SSIM, RMSE, FID, and NCC indicator values on four sub-datasets and the SIMLE-CR dataset are better than the seven comparison methods. Furthermore, the ablation experiments show that the generalization and robustness of this proposed method on images with different cloud coverage are better than other comparison methods. Therefore, DG2-TCR can reliably recover information on cloud occlusions with various coverage and thickness, which is significant for cloud removal in practical applications.

The global cloud detection of high-resolution remote sensing images (HRSI) is crucial for acquiring high-quality imagery and optimizing data utilization. Traditional cloud detection models, which rely on limited samples and fully supervised learning, struggle to adapt to cross-temporal and cross-spatial domains. While current unsupervised domain adaptation (UDA) methods improve performance in cross-domain cloud detection to some extent, generating high-quality, reliable pseudo-labels remains a significant challenge for global cloud detection. Therefore, this paper proposes a progressive pseudo-label learning for cross-domain cloud detection in high-resolution remote sensing (PTDA). Firstly, we propose an online domain-invariant feature guided pseudo-label generation (OPLG) strategy and learning intra-domain unaligned features (LIUF), which effectively integrate domain-invariant features and intra-domain semantics to generate high-quality pseudo-labels at the feature level. LIUF then refines the pseudo-label quality at the pixel level. Secondly, during the model training, pseudo-label constrained intra-domain feature mining loss(PCIF Loss) is designed to suppress noisy semantic information within the domain, the hole effect of thick/thin clouds, and the noise interference of the contour boundary. Four cloud detection datasets, including MS Cloud (MS), HRC WHU Cloud (WHU), 95 Cloud(95), and WHUS2-CD+(S2), are grouped into three cross-domain tests, MS2WHU, MS2S2, and WHU295. Our approach achieved the best performance with mIoU 63.99%, 58.14%, 58.82%, and OA 80.03%, 79.36%, 80.49%, respectively. The experimental results show that the proposed method outperforms seven state-of-art cross-domain comparison methods. Thus, our method has important application value for cross-domain cloud detection. The available code can be downloaded from https://github.com/gasking/PTDA.