Using images captured by cameras with different light spectrum sensitivities, training a unified model for cross-spectral scene representation is challenging. Recent advances have shown the possibility of jointly optimizing cross-spectral relative poses and neural radiance fields using normalized cross-device coordinates. However, such method suffers from cross-spectral misalignment when collecting data asynchronously from devices and lacks the capability to render in real-time or handle large scenes. We address these issues by proposing cross-spectral Gaussian Splatting with spatial occupancy consistency, strictly aligns cross-spectral scene representation by sharing explicit Gaussian surfaces across spectra and separately optimizing each view's extrinsic using a matching-optimizing pose estimation method. Additionally, to address field-of-view differences in cross-spectral cameras, we improve the adaptive densify controller to fill non-overlapping areas. Comprehensive experiments demonstrate that SOC-GS achieves superior performance in novel view synthesis and real-time cross-spectral rendering.

A discernible gap has materialized between the expectations for object detection tasks in optical remote sensing images and the increasingly sophisticated design methods. The flexibility of deep learning object detection algorithms allows the selection and combination of multiple basic structures and model sizes, but this selection process relies heavily on human experience and lacks reliability when faced with special scenarios or extreme data distribution. To address these inherent challenges, this study proposes an approach that leverages deep reinforcement learning within the framework of vision tasks. This study introduces a Task-Risk Consistent Intelligent Detection Framework (TRC-ODF) for object detection in optical remote sensing images. The proposed framework designs a model optimization strategy based on deep reinforcement learning that systematically integrates the available information from images and vision processes. The core of the reinforcement learning agent is the proposed task-risk consistency reward mechanism, which is the driving force behind the optimal prediction allocation in the decision-making process. To verify the effectiveness of the proposed framework, multiple sets of empirical evaluations are conducted on representative optical remote sensing image datasets: RSOD, NWPU VHR-10, and DIOR. When applying the proposed framework to representative advanced detection models, the mean average precision (mAP@0.5 and mAP@0.5:0.95) is improved by 0.8–5.4 and 0.4–2.7, respectively. The obtained results showcase the considerable promise and potential of the TRC-ODF framework to address the challenges associated with object detection in optical remote sensing images.

Person search has advanced rapidly with end-to-end frameworks that jointly integrate person detection and re-identification. However, retrieval quality remains highly query dependent: under a fixed gallery and evaluation protocol, a single fixed model often performs inconsistently across queries under heterogeneous visual conditions. To address this issue, we propose ReQAR, a model-agnostic query adaptive model routing framework for person search. Instead of relying on a single fixed model, ReQAR selects the most suitable model for each query from a pool of complementary end-to-end person search models at inference time, without modifying model architectures or weights. To support reliable routing, ReQAR employs a lightweight query-state encoder, ResQENet, to produce compact and decision-relevant query representations. The routing policy is learned with retrieval-metric-aligned rewards, enabling effective exploitation of complementary model strengths. Extensive experiments on CUHK-SYSU and PRW show that ReQAR consistently outperforms the strongest single-model baseline in the candidate pool, achieving 97.0% mAP / 97.5% Top-1 on CUHK-SYSU and 57.6% mAP / 90.3% Top-1 on PRW. These results suggest that query-adaptive model routing provides an effective alternative to purely model-centric improvement for end-to-end person search under diverse query conditions.