Low-lightimage enhancement methods are difficult to fit the complex nonlinear relationship between normal and low-light images due to uneven illumination and noise effects. The recently proposed Kolmogorov-Arnold networks (KANs) feature spline-based convolutional layers and learnable activation functions, which can effectively capture nonlinear dependencies. In this paper, we design a KAN-Block based on KANs and innovatively apply it to low-light image enhancement. This method effectively alleviates the limitations of current methods constrained by linear network structures and lack of interpretability, further demonstrating the potential of KANs in low-level vision tasks. Given the poor perception of current low-light image enhancement methods and the stochastic nature of the inverse diffusion process, we further introduce frequency-domain perception for visually oriented enhancement. Extensive experiments demonstrate the competitive performance of our method on benchmark datasets.

In semi-supervised medical image segmentation, the issue of fuzzy boundaries for segmented objects arises. With limited labeled data and the interaction of boundaries from different segmented objects, classifying segmentation boundaries becomes challenging. To mitigate this issue, we propose an uncertainty global contrastive learning (UGCL) framework. Specifically, we propose a patch filtering method and a classification entropy filtering method to provide reliable pseudo-labels for unlabelled data, while separating fuzzy boundaries and high-entropy pixel points as unreliable points. Considering that unreliable regions contain rich complementary information, we introduce an uncertainty global contrast learning method to distinguish these challenging unreliable regions, enhancing intra-class compactness and inter-class separability at the global data level. Within our optimization framework, we also integrate consistency regularization techniques and select unreliable points as targets for consistency. As demonstrated, the contrastive learning and consistency regularization applied to uncertain points enable us to glean valuable semantic information from unreliable data, which enhances segmentation accuracy. We evaluate our method on two publicly available medical image datasets and compare it with other state-of-the-art semi-supervised medical image segmentation methods, and a series of experimental results show that our method has achieved substantial improvements.

Graph-based clustering performs efficiently for identifying clusters in local and nonlinear data Patterns. The existing methods face the problem of parameter selection, such as the setting of k of the k-nearest neighbor graph and the threshold in noise detection. In this paper, a non-parametric clustering algorithm (NonPC) is proposed to tackle those inherent limitations and improve clustering performance. The weighted natural neighbor graph (wNaNG) is developed to represent the given data without any prior knowledge. What is more, the proposed NonPC method adaptively detects noise data in an unsupervised way based on some attributes extracted from wNaNG. The algorithm works without preliminary parameter settings while automatically identifying clusters with unbalanced densities, arbitrary shapes, and noises. To assess the advantages of the NonPC algorithm, extensive experiments have been conducted compared with some classic and recent clustering methods. The results demonstrate that the proposed NonPC algorithm significantly outperforms the state-of-the-art and well-known algorithms in Adjusted Rand index, Normalized Mutual Information, and Fowlkes-Mallows index aspects.

It is necessary to detect the license plate and the parking space number in autonomous driving. Due to the complex environment of the parking space, it is challenging and interesting in parking space number detection. This paper proposes a robust Parking Space Number Detector (PSND), which enhances the Differentiable Binarization (DB) model by introducing a cascaded feature enhancement module and context attention block. Our method has better feature extraction ability and a better detection effect for long text than the DB model. Meanwhile, we collected and annotated a dataset containing 9000 parking space number images to train and test our model. Extensive experiments demonstrate that our method achieves better or competitive performance in terms of accuracy on various standard benchmarks, including MSRA-TD500, ICDAR2015, CTW1500, and Total-Text datasets while maintaining the real-time detection speed. The high recognition accuracy makes it possible for the model to be practically applied.

Stacking can improve the signal-to-noise ratio (SNR) of seismic data, and therefore, it plays an important role in seismic signal processing. The commonly used stacking methods involve calculating a weighted average trace to reduce the influence of harmful samples with random noise and imprecise travel-time correction. However, abnormal misaligned traces often cause large amplitude change in the same reflection time, leading to very small or even zero weights for the purpose of improving SNR. Therefore, the existed weighted stacking methods discard too much seismic reflection information and may not be conducive to stacking. In this letter, we propose a novel weighted stacking method that uses dynamic time warping (DTW) to solve the misalignment problem of seismic reflection events. Then, the weighted stacking can be applied to suppress random noise based on the aligned samples. This proposed approach will enhance the role of the misaligned traces using larger weights without introducing additional noise, thus making the stacking procedure more robust. The application results on both synthetic and real seismic data demonstrate the effectiveness of our proposed approach.

Delineating the seismic fault, which is an important type of geologic structures in seismic images, is a key step for seismic interpretation. Comparing with conventional methods that design a number of hand-crafted features based on the observed characteristics of the seismic fault, convolutional neural networks (CNNs) have proven to be more powerful for automatically learning effective representations. However, the CNN usually serves as a black box in the process of training and inference, which would lead to trust issues. The inability of humans to understand the CNN would be more problematic, especially in critical areas like seismic exploration, medicine and financial markets. To include domain knowledge to improve the interpretability of the CNN, we propose to jointly optimize the prediction accuracy and consistency between explanations of the neural network and domain knowledge. Taking the seismic fault segmentation as an example, we show that the proposed method not only gives reasonable explanations for its predictions, but also more accurately predicts faults than the baseline model.

One main objective of Soil Moisture and Ocean Salinity (SMOS) is to provide observations of global soil moisture. The resolution of L-band radiometer carried by SMOS is about 50 km resolution, which is rather coarse for many applications. The method based on `universal triangle' has been used on downscaling soil moisture retrieved by SMOS with L-band brightness temperature. Consider the senor carried by SMOS is the L-band multi-angular dual polarization radiometer, this paper presents an improved method for downscaling Centre Aval de Traitement des Données SMOS (CATDS) L3 soil moisture by using MODIS Land Surface Temperature (LST) /NDVI and SMOS dual polarization brightness temperature. Southwest of China is selected as the study area and ground based soil moisture from Maqu monitor network is used to compare with the SMOS CATDS soil moisture at 25 km and the downscaled soil moisture. The results show that the improved method could improve the result with R2from 0.15 to 0.25. High resolution brightness temperature used in the method is obtained from SMOS CATDS brightness temperature at 25 km and causes error during the experiment.

This paper presents a method to retrieve surface soil moisture at high-resolution from brightness temperature. To obtain high resolution brightness temperature, downscaling brightness temperature uses higher resolution visible/infrared satellite data and is based on the method of retrieving land surface temperature with passive microwave and the relationship between Microwave Polarization Difference Index (MPDI) and NDVL High resolution soil moisture is retrieved with downscaled brightness temperature using a single-channel algorithm (SCA), and the Qpmodel used to deal with the influence of roughness. Downscaled brightness temperature is compared with the aircraft data during the Experiment in the Heihe River Basin in China. The retrieved high resolution soil moisture is compared with in situ (0-6 cm depth) point measurements during SMEX02, obtaining R2= 0.74 and RMSE as 0.047 cm3cm-3.