As a fundamental preprocessing technique, remote sensing image motion deblurring is important for visual understanding tasks. Most conventional approaches formulate the image motion deblurring task as a kernel estimation. Because the kernel estimation is a highly ill-posed problem, many priors have been applied to model the images and kernels. Even though these methods have obtained relatively better performances, they are usually time-consuming and not robust for different conditions. To address this problem, we propose a multi-task learning mechanism for remote sensing image motion deblurring in this paper, which contains an image restoration subtask and an image texture complexity recognition one. First, we consider the image motion deblurring problem as a domain transformation problem, from the blurred domain to a clear one. Specifically, the blurred domain represents the data space consisted of blurring images, and the defination of clear domain is similar. Second, we design a novel weighted attention map loss to enhance the reconstruction capability of the restoration subbranch for difficult local regions. Third, based on the restoration subbranch, a recognition subbranch is incorporated into the framework to guide the deblurring process, which provides the auxiliary texture complexity information to help the optimization of restoration subbranch. Additionally, in order to optimize the proposed network, we construct three large-scale datasets, and each sample in the dataset contains a clear image, a blurred image and its texture label obtained by corresponding texture complexity. Finally, the experimental results on three constructed datasets demonstrate the robustness and the effectiveness of the proposed method.

Depth estimation in low-light scenes is an enormous challenge in the field of monocular depth estimation. Although numerous algorithms have attempted to improve their performance in low-light scenarios through a variety of techniques, the inconsistent illumination issue caused by local intense or colored light is rarely taken into consideration. To tackle this problem, we proposed an illumination adaptive image enhancement-based low-light depth estimation algorithm (IAIE_LDE) in this paper. Our main contribution is an S-shaped illumination estimation basis illumination adaptive consistent correction model, which is designed to eliminate the edge blurring and depth hole effects in depth maps caused by inconsistent lighting. Meanwhile, a low-light depth estimation architecture consisting of three modules, namely, illumination adaptive correction, low-light image enhancement and depth estimation modules, is constructed and trained. Specifically, the first sub-module is designed to alleviate the illumination inconsistency utilizing the proposed S-shaped illumination adaptive correction model by calculating the corresponding correction coefficients for each pixel according to the estimated illumination; the core module of the classic EnlightGAN algorithm is adopted in the second sub-module to improve the overall brightness of the image and solve the other problems caused by low light; the ZoeDepth model is chosen as our depth estimation sub-module to output a depth map comparable to high-quality illuminated images. Extensive experiments on the widely used Oxford RobotCar and nuScenes datasets indicate superior performance of our method by comparing it with state-of-the-art low-light depth estimation algorithms such as RNW, STEPS, ADDS-DepthNet, and ACDepth, both qualitatively and quantitatively.

Achieving favorable compatibility between detection performance and inference efficiency has long been a core bottleneck restricting the practical deployment of vision-based driver fatigue detection algorithms on vehicle-mounted embedded devices. In this work, we propose an adaptive temporal-aware feature selection paradigm for driver fatigue detection to address temporal isolation and computational redundancy problems. Specifically, we propose a temporal dual-stream fusion (TDF) module to generate unified embeddings of highly discriminative video features, laying a solid feature foundation for superior final performance. Additionally, we design an adaptive feature selection (AFS) mechanism that evaluates the necessity of executing each stream based on confidence estimation, eliminating redundant computations with acceptable slight accuracy loss. In addition, the experimental results on the challenging NTHU-DDD dataset demonstrate the superiority of our method, especially from the perspective of balancing performance and efficiency.

To tackle the challenging representation problem of fine-grained tire tread patterns, we present a hierarchical sparse-routing-enabled multi-mode multi-directional Mamba (HSR-Mamba) scanning framework for tire pattern recognition to improve performance, and it mainly consists of a fine-grained textural MoE based on four-directional pixel scanning and a coarse-grained structural Mixture-of-Experts (MoE) based on three global patch traversals. It simultaneously realizes micro multi-directional fine-grained texture modeling and macro multi-mode global structure modeling. Additionally, hierarchical sparse routing is adopted to eliminate redundant forward computations, which effectively balances recognition accuracy and inference efficiency through correlated activation mechanism of a dual-layer expert system. In addition, a series of experimental results on CIIP-TPID-V1.1 have verified its effectiveness.