Adjacent Features Research Articles

Cross-Layer Triple-Branch Parallel Fusion Network for Small Object Detection in UAV Images

Detection of small objects in Unmanned Aerial Vehicle (UAV) images, which contain a large number of objects with extremely small pixels, has been a major challenge over the last few decades. Owing to the lack of sufficiently detailed representation of object features, conventional detectors based on deep learning are unsatisfactory when detecting small objects. Meanwhile, the flight altitude and the shooting angle of the drone always changing, which leads to an uneven dispersion of the multiple objects and a varying density of small objects. In response to the issues mentioned above, a novel lightweight Cross-layer Triple-branch Parallel Fusion Network (CTPFNet) is proposed to improve the real-time detection accuracy of small objects under UAV images. Firstly, a novel downsampling structure called the Inverse Residual Pooling Cascade (IRPC) module is proposed to obtain richer feature information about small objects. Secondly, we construct an improved global feature extraction structure Efficient Layer Aggregation Networks-Transformer (ELAN-Trans) to enhance the association between global features. Then, we design the Hybrid Dilated Depthwise Separable-Spatial Pyramid Pooling-Fast (HDDS-SPPF) to capture more contextual information by using the dilated convolution operations. To enhance the cross-scale transfer fusion of low-level original features containing more detailed localization information and deep-level deep features containing more semantic information, we propose a CTPF module embedded into the neck region for the secondary feature reuse of adjacent feature maps with varying resolutions. Extensive experiments on the public VisDrone2021-DET dataset show that the proposed model achieves significant performance gains with fewer parameters.

Image Manipulation Localization Using Multi-Scale Feature Fusion and Adaptive Edge Supervision

Image manipulation localization is a technique that can efficiently segment the tampered regions from a suspicious image. Existing work usually trains a detection model by fusing the features from diverse data streams, e.g., noise inconsistency, recompression inconsistency, and local inconsistency. They, however, ignore a fact that not all tampered images contain these data streams. As a result, high feature redundancy may cause a large number of false detection for tampered region. To address this problem, this paper designs an end-to-end high-confidence localization network architecture. First, deep convolutional neural networks are utilized to extract multi-scale feature sets from the RGB streams. We then design a semantic refined bi-directional feature integration module to fully fuse multi-scale adjacent features and significantly enhance feature representation. Subsequently, morphological operations are introduced to extract multi-scale edge information, which can efficiently reduce feature redundancy by generating wider high-resolution edges during image reconstructing. Finally, a deep semantic residual decoder is sequentially re-constructed by spreading deep semantic information into each decoding stage. The proposed method can not only improve the manipulation localization accuracy, but also guarantee the model robustness. Extensive experiments demonstrate that our method can obtain an effective performance in locating forged regions over different large-scale image sets, and outperforms most of state-of-the-art methods with higher localization accuracy and stronger robustness.

ShapeFormer: A Shape-Enhanced Vision Transformer Model for Optical Remote Sensing Image Landslide Detection

Landslides pose a serious threat to human life, safety, and natural resources. Remote sensing images can be used to effectively monitor landslides at a large scale, which is of great significance for pre-disaster warning and post-disaster assistance. In recent years, deep learning based methods have made great progress in the field of remote sensing image landslide detection. In remote sensing images, landslides display a variety of scales and shapes. In this paper, to better extract and keep the multi-scale shape information of landslides, a shape-enhanced vision transformer (ShapeFormer) model is proposed. For the feature extraction, a pyramid vision transformer (PVT) model is introduced, which directly models the global information of local elements at different scales. To learn the shape information of different landslides, a shape feature extraction branch is designed, which uses the adjacent feature maps at different scales in the PVT model to improve the boundary information. After the feature extraction step, a decoder with deconvolutional layers follows, which combines the multiple features and gradually recovers the original resolution of the combined features. A softmax layer is connected with the combined features to acquire the final pixel-wise result. The proposed ShapeFormer model was tested on two public datasets—the Bijie dataset and the Nepal dataset—which have different spectral and spatial characteristics. The results, when compared with those of some of the state-of-the-art methods, show the potential of the proposed method for use with multi-source optical remote sensing data for landslide detection.

Smoking Detection Algorithm Based on Improved YOLOv5

Aiming at the current problems in smoking detection in public places, such as many small and medium-sized cigarette targets, low pixels occupied by cigarette targets, indistinct characteristics of cigarette targets, and difficult detection, an improved smoking detection algorithm YOLOv5_EC is proposed. Based on the YOLOv5, this paper redesigns the neck structure and proposes the EFFN (Enhanced Feature Fusion Neck) structure, which fuses three adjacent feature maps of different sizes to better retain the positioning information of the target, further improves the feature expression ability of small objects, and then introduce the CBAM attention module in the network, so that the model can focus on the important areas of the image, suppress the interference of irrelevant information, enhances the model's ability to learn features, and improves the accuracy of model detection. Experiments show that the mAP@0.5 of the model proposed in this paper is improved by 1.5% compared with the original YOLOv5 model, and the improved model can effectively identify smoking behavior in actual scenes.

Adjacent Context Coordination Network for Salient Object Detection in Optical Remote Sensing Images.

Salient object detection (SOD) in optical remote sensing images (RSIs), or RSI-SOD, is an emerging topic in understanding optical RSIs. However, due to the difference between optical RSIs and natural scene images (NSIs), directly applying NSI-SOD methods to optical RSIs fails to achieve satisfactory results. In this article, we propose a novel adjacent context coordination network (ACCoNet) to explore the coordination of adjacent features in an encoder-decoder architecture for RSI-SOD. Specifically, ACCoNet consists of three parts: 1) an encoder; 2) adjacent context coordination modules (ACCoMs); and 3) a decoder. As the key component of ACCoNet, ACCoM activates the salient regions of output features of the encoder and transmits them to the decoder. ACCoM contains a local branch and two adjacent branches to coordinate the multilevel features simultaneously. The local branch highlights the salient regions in an adaptive way, while the adjacent branches introduce global information of adjacent levels to enhance salient regions. In addition, to extend the capabilities of the classic decoder block (i.e., several cascaded convolutional layers), we extend it with two bifurcations and propose a bifurcation-aggregation block (BAB) to capture the contextual information in the decoder. Extensive experiments on two benchmark datasets demonstrate that the proposed ACCoNet outperforms 22 state-of-the-art methods under nine evaluation metrics, and runs up to 81 fps on a single NVIDIA Titan X GPU. The code and results of our method are available at https://github.com/MathLee/ACCoNet.

DAMNet: A Dual Adjacent Indexing and Multi-Deraining Network for Real-Time Image Deraining

Image deraining is increasingly critical in the domain of computer vision. However, there is a lack of fast deraining algorithms for multiple images without temporal and spatial features. To fill this gap, an efficient image-deraining algorithm based on dual adjacent indexing and multi-deraining layers is proposed to increase deraining efficiency. The deraining operation is based on two proposals: the dual adjacent method and the joint training method based on multi-deraining layers. The dual adjacent structure indexes pixels from adjacent features of the previous layer to merge with features produced by deraining layers, and the merged features are reshaped to prepare for the loss computation. Joint training method is based on multi-deraining layers, which utilise the pixelshuffle operation to prepare various deraining features for the multi-loss functions. Multi-loss functions jointly compute the structural similarity by loss calculation based on reshaped and deraining features. The features produced by the four deraining layers are concatenated in the channel dimension to obtain the total structural similarity and mean square error. During the experiments, the proposed deraining model is relatively efficient in primary rain datasets, reaching more than 200 fps, and maintains relatively impressive results in single and crossing datasets, demonstrating that our deraining model reaches one of the most advanced ranks in the domain of rain-removing.

Unsupervised medical image feature learning by using de-melting reduction auto-encoder

Unsupervised feature learning is a fundamental and highly prioritized problem in medical image analysis. Although it has shown considerable improvements, it remains challenging because of its weak feature expression ability, low model-learning efficiency, and weak robustness. To address these limitations, a novel unsupervised feature learning method in the medical image classification task, named de-melting reduction auto-encoder (DMRAE), is proposed in this study. A joint fusion network structure is constructed; it not only improves the expression of target features but also reduces the loss of feature decoding and parameters. To obtain a robust solution, a newly designed decomposed-reconstructed loss function is used to strengthen the semantic context between adjacent feature extractor layers, successfully avoiding the insufficient model-learning ability from the single optimization objective and improving the quality of the extracted features. Finally, extensive experiments on datasets consisting of 400 breast ultrasonographic images and 6000 lung computed tomography images are conducted to demonstrate the effectiveness of the proposed method. Experimental results reveal that the DMRAE significantly reduces the annotation effort and outperforms existing methods by a significant margin.

Improving multispectral pedestrian detection with scale‐aware permutation attention and adjacent feature aggregation

AbstractHigh quality feature fusion module is one of the key components for multispectral pedestrian detection system in challenging situations, such as large‐scale variance and occlusion. Although attention mechanism is one of the most effective ways for feature refining, the correlation between attention and scales in feature pyramid still remains unknown. Therefore, a scale‐aware permutated attention module is proposed to enhance features of objects with different scales adaptively in the feature pyramid. Specifically, four different local and global attention sub‐modules are investigated to refine feature maps with different permutations in the Feature Pyramid Networks, improving the quality of the feature fusion. Besides, to address the high miss‐rate issue for small‐sized pedestrians, an adjacent‐branch feature aggregation module is proposed to aggregate features across different scales, taking both semantic context and spatial resolution into consideration. Both modules can benefit from each other with significant performance improvement in terms of efficiency and accuracy, when equipped with the dual‐branch CenterNet detection framework. Experiments on the KAIST and FLIR datasets demonstrate its superior performance compared with other state‐of‐the‐arts.

Short‐term load forecasting of multi‐scale recurrent neural networks based on residual structure

SummaryAccurate short‐term load forecasting plays an important role in reducing power generation costs, maintaining supply and demand balance, and stabling the power grids operation. In recent years, deep learning models based on recurrent neural networks (RNN) have been widely used in short‐term load forecasting. Nevertheless, RNN cannot extract multi‐scale features of load data, resulting in low forecasting accuracy. A model for short‐term power load forecasting of residual multiscale‐RNN (RM‐RNN) was proposed in this study. RM‐RNN uses the multilayer RNN network structure. Specifically, each layer sets the dilated convolution with different dilated coefficients to extract the multi‐scale features of the load data. Adjacent networks transfer feature information for feature fusion through the residual structures. The experiment used random sampling data training model, and compared RM‐RNN with multiple deep learning models. The experimental results demonstrated that the mean error of RM‐RNN prediction is the lowest, indicating that dilated convolution can effectively extract multi‐scale features of load data. This result verified the effectiveness of residual structure fusion features, and improved the accuracy of short‐term load forecasting.

VGG-TSwinformer: Transformer-based deep learning model for early Alzheimer’s disease prediction

Background and objective: Mild cognitive impairment (MCI) is a transitional state between normal aging and Alzheimer’s disease (AD), and accurately predicting the progression trend of MCI is critical to the early prevention and treatment of AD. Brain structural magnetic resonance imaging (sMRI), as one of the most important biomarkers for the diagnosis of AD, has been applied in various deep learning models. However, due to the inherent disadvantage of deep learning in dealing with longitudinal medical image data, few applications of deep learning for longitudinal analysis of MCI, and the majority of existing deep learning algorithms for MCI progress prediction rely on the analysis of the sMRI images collected at a single time-point, ignoring the progressive nature of the disorder. Methods: In this work, we propose a VGG-TSwinformer model based on convolutional neural network (CNN) and Transformer for short-term longitudinal study of MCI. In this model, VGG-16 based CNN is used to extract low-level spatial features of longitudinal sMRI images and map these low-level features to high-level feature representations, sliding-window attention is used for fine-grained fusion of spatially adjacent feature representations, and gradually fuses distant spatial feature representations through the superposition of attention windows of different sizes, temporal attention is used to measure the evolution of this feature representations as a result of disease progression. Results: We validated our model on the ADNI dataset. For the classification task of sMCI vs pMCI, accuracy, sensitivity, specificity and AUC reached 77.2%, 79.97%, 71.59% and 0.8153 respectively. Compared with other cross-sectional studies also applied to sMRI, the proposed model achieved better results in terms of accuracy, sensitivity, and AUC. Conclusion: The proposed VGG-TSwinformer is a deep learning model for short-term longitudinal study of MCI, which can build brain atrophy progression model from longitudinal sMRI images, and improve diagnostic efficiency compared to algorithms using only cross-sectional sMRI images.

Scale Enhancement Pyramid Network for Small Object Detection from UAV Images.

Object detection is challenging in large-scale images captured by unmanned aerial vehicles (UAVs), especially when detecting small objects with significant scale variation. Most solutions employ the fusion of different scale features by building multi-scale feature pyramids to ensure that the detail and semantic information are abundant. Although feature fusion benefits object detection, it still requires the long-range dependencies information necessary for small objects with significant scale variation detection. We propose a simple yet effective scale enhancement pyramid network (SEPNet) to address these problems. A SEPNet consists of a context enhancement module (CEM) and feature alignment module (FAM). Technically, the CEM combines multi-scale atrous convolution and multi-branch grouped convolution to model global relationships. Additionally, it enhances object feature representation, preventing features with lost spatial information from flowing into the feature pyramid network (FPN). The FAM adaptively learns offsets of pixels to preserve feature consistency. The FAM aims to adjust the location of sampling points in the convolutional kernel, effectively alleviating information conflict caused by the fusion of adjacent features. Results indicate that the SEPNet achieves an AP score of 18.9% on VisDrone, which is 7.1% higher than the AP score of state-of-the-art detectors RetinaNet achieves an AP score of 81.5% on PASCAL VOC.

Novel Asymmetric Pyramid Aggregation Network for Infrared Dim and Small Target Detection

Robust and efficient detection of small infrared target is a critical and challenging task in infrared search and tracking applications. The size of the small infrared targets is relatively tiny compared to the ordinary targets, and the sizes and appearances of the these targets in different scenarios are quite different. Besides, these targets are easily submerged in various background noise. To tackle the aforementioned challenges, a novel asymmetric pyramid aggregation network (APANet) is proposed. Specifically, a pyramid structure integrating dual attention and dense connection is firstly constructed, which can not only generate attention-refined multi-scale features in different layers, but also preserve the primitive features of infrared small targets among multi-scale features. Then, the adjacent cross-scale features in these multi-scale information are sequentially modulated through pair-wise asymmetric combination. This mutual dynamic modulation can continuously exchange heterogeneous cross-scale information along the layer-wise aggregation path until an inverted pyramid is generated. In this way, the semantic features of lower-level network are enriched by incorporating local focus from higher-level network while the detail features of high-level network are refined by embedding point-wise focus from lower-level network, which can highlight small target features and suppress background interference. Subsequently, recursive asymmetric fusion is designed to further dynamically modulate and aggregate high resolution features of different layers in the inverted pyramid, which can also enhance the local high response of small target. Finally, a series of comparative experiments are conducted on two public datasets, and the experimental results show that the APANet can more accurately detect small targets compared to some state-of-the-art methods.

MB-Net: multiscale boundary interaction learning for image manipulation localization

Image editing techniques can modify the content of images indiscriminately, which causes a grave threat to the security of society. Hence, the localization of manipulated images is inevitable. A serious challenge for image manipulation detection is the lack of strategies for perceiving global features and refining edges. In this paper, we present a multiscale boundary interaction learning network for image manipulation localization to solve both problems. This network contains an adjacent-scale mutual module to enrich the global perception domain by interactively learning adjacent scale features. It avoids the tremendous noise interference caused by the direct fusion of all scale features. To effectively suppress semantic content segmentation, the boundary pixel disparity module computes interpixel differences at specific angles to enhance boundary artifact recognition between tampered and real regions. The fusion attention module is proposed to combine scale and edge messages, integrating spatial and channel correlations in a compatible way. Extensive experimental results indicate that our proposed method is significantly superior to current state-of-the-art methods on public standard datasets.

Deep-Separation Guided Progressive Reconstruction Network for Semantic Segmentation of Remote Sensing Images

The success of deep learning and the segmentation of remote sensing images (RSIs) has improved semantic segmentation in recent years. However, existing RSI segmentation methods have two inherent problems: (1) detecting objects of various scales in RSIs of complex scenes is challenging, and (2) feature reconstruction for accurate segmentation is difficult. To solve these problems, we propose a deep-separation-guided progressive reconstruction network that achieves accurate RSI segmentation. First, we design a decoder comprising progressive reconstruction blocks capturing detailed features at various resolutions through multi-scale features obtained from various receptive fields to preserve accuracy during reconstruction. Subsequently, we propose a deep separation module that distinguishes various classes based on semantic features to use deep features to detect objects of different scales. Moreover, adjacent middle features are complemented during decoding to improve the segmentation performance. Extensive experimental results on two optical RSI datasets show that the proposed network outperforms 11 state-of-the-art methods.

Deep Forest With Multi-Channel Message Passing and Neighborhood Aggregation Mechanisms for Brain Network Classification.

As a novel deep learning method, deep forest has achieved excellent classification performance on many small-scale datasets, thus providing a new opportunity to accurately classify brain networks (BNs) on limited fMRI data. Though there are a few explorations about classifying BNs using deep forest, they only adopt sliding windows to extract adjacent features of BNs and fail to use prior knowledge to strengthen the features more relevant to brain diseases. In this paper, we propose a deep forest framework with multi-channel message passing and neighborhood aggregation mechanisms (DF-MCMPNA) to extract and aggregate long-range multi-channel topological features. Firstly, we use the three intrinsic connectivity networks (ICNs) and the whole-brain to form four feature extraction channels. Secondly, we present a multi-channel message passing mechanism and a channel-shared neighborhood aggregation mechanism to recursively extract long-range multi-channel topological features, where the first mechanism can learn local topological features in each channel and the second mechanism can fuse multi-channel topological features. Finally, the extracted features are fed into the casForst to perform further feature learning and classification. Experimental results on ABIDE I, ABIDE II, and ADHD-200 datasets show that the DF-MCMPNA outperforms several state-of-the-art methods on classification performance and accurately identifies abnormal brain regions.

Progressive Multiscale Consistent Network for Multiclass Fundus Lesion Segmentation.

Effectively integrating multi-scale information is of considerable significance for the challenging multi-class segmentation of fundus lesions because different lesions vary significantly in scales and shapes. Several methods have been proposed to successfully handle the multi-scale object segmentation. However, two issues are not considered in previous studies. The first is the lack of interaction between adjacent feature levels, and this will lead to the deviation of high-level features from low-level features and the loss of detailed cues. The second is the conflict between the low-level and high-level features, this occurs because they learn different scales of features, thereby confusing the model and decreasing the accuracy of the final prediction. In this paper, we propose a progressive multi-scale consistent network (PMCNet) that integrates the proposed progressive feature fusion (PFF) block and dynamic attention block (DAB) to address the aforementioned issues. Specifically, PFF block progressively integrates multi-scale features from adjacent encoding layers, facilitating feature learning of each layer by aggregating fine-grained details and high-level semantics. As features at different scales should be consistent, DAB is designed to dynamically learn the attentive cues from the fused features at different scales, thus aiming to smooth the essential conflicts existing in multi-scale features. The two proposed PFF and DAB blocks can be integrated with the off-the-shelf backbone networks to address the two issues of multi-scale and feature inconsistency in the multi-class segmentation of fundus lesions, which will produce better feature representation in the feature space. Experimental results on three public datasets indicate that the proposed method is more effective than recent state-of-the-art methods.

Traffic patterns, more than adjacent land use, influence element content of roadside forbs for insect pollinators

Abstract Roadsides are targeted for restoration of pollinator‐friendly plants. Yet, roads are sources of macronutrient, micronutrient and heavy metal pollution that may contaminate roadside plants. Adjacent landscape features such as railroads and agriculture provide additional macronutrient and heavy metal pollution that may exacerbate traffic effects. However, we lack perspective on how roads combine with rural landscape features to influence nutrition of roadside plants, which could have implications for pollinator health. We surveyed roadsides across Minnesota, USA and measured foliar levels of dietary macronutrients (nitrogen, phosphorous and potassium), a micronutrient (sodium) and metals (iron, zinc, copper, chromium, nickel, lead, aluminium and cadmium) in six abundant roadside forb species used by insect pollinators: Asclepias syriaca, Dalea purpurea, Monarda fistulosa, Ratibida pinnata, Solidago spp. and Trifolium pratense. We aimed to determine (1) how road variables (traffic volume and distance from road) combine with adjacent land use (railroad and agriculture) to influence element content of roadside forbs and (2) whether some forb species show consistent differences in their accumulation of potentially toxic heavy metals, which could inform selection of species to plant along roadsides. We found that foliar concentrations of nine elements increased with greater traffic volume (nitrogen, phosphorous, iron, zinc, copper, chromium, nickel, lead and aluminium), and concentrations of six elements declined with distance from the road (nitrogen, phosphorous, potassium, iron, zinc and copper). Leaves collected adjacent to railroad had less phosphorous, potassium, iron, nickel and aluminium than leaves collected from sites not adjacent to railroad. Additionally, leaves collected from sites adjacent to agriculture had lower copper levels than leaves from sites without adjacent agriculture. We found no evidence that particular ford species along roadsides consistently rank higher than other species in their accumulation of heavy metals. Our results show that traffic alters more elements in roadside plants than does adjacent landscape context, alleviating concerns that landscape features exacerbate pollutant levels in roadside pollinator habitat. However, nutrient contamination of most roadside plants is unlikely to reach toxic levels for insect pollinators. This work is consistent with the positive conservation potential of low to moderate traffic roadsides for pollinators.

Foreign Body Detection in Rail Transit Based on a Multi-Mode Feature-Enhanced Convolutional Neural Network

Detection of railway traffic objects is an important task during train driving and is implemented to ensure safe driving. Although object detection has been investigated for years, many challenges exist in precisely detecting railway objects under complex railway scenes. These challenges mainly include adverse weather states, various railway backgrounds, diverse railway objects, and low-quality images. To address these issues, we introduce a novel deep learning method, called a multi-mode feature enhanced convolutional neural network (MMFE-Net), for accurate railway object detection. The network mainly consists of three modules. 1) An improved cross-stage partial connection darknet53 (CSPDarknet53), called adaptive dilated cspdarknet53, is used as our backbone to reduce image information loss. 2) A spatial feature extraction module is used to improve the feature extraction ability of the model for blurred objects and objects in a complicated background. 3) We introduce an attention fusion enhance module to strengthen the context information between adjacent feature maps to accurately detect multiscale and small objects. The proposed method achieves 0.9439 mAP and 79 FPS with an input size of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$640\times 640$ </tex-math></inline-formula> pixels on the railway traffic dataset, and its performance is better than that of YOLOv4. Moreover, it is feasible to apply MMFE-Net into practical applications of railway object detection.

Deep learning‐based super‐resolution with feature coordinators preservation for vision‐based measurement

Vision-based displacement measurement is promising and emerging for structural monitoring. However, the accuracy of visual measurement is commonly limited by the resolution of the camera. The super-resolution (SR) technique can reconstruct high-resolution images from the corresponding low-resolution images within the constraints of prior knowledge. Existing SR algorithms mainly focus on improving the overall quality of the image. By contrast, the accurate extraction of the coordinates of feature points is the most important for the visual measurement. Besides, the SR network is usually trained by an artificial dataset whose low-resolution images are obtained by artificially degrading the corresponding high-resolution images, instead of those directly captured by cameras. However, this degradation usually is only a simple bicubic downsampling that cannot reflect the real degradation, which will provide inaccurate constraints to the network training. Therefore, this paper proposes a novel SR framework that can significantly preserve the feature coordinators for visual measurement (SRFCP). First, a deep learning-based SR network that focuses on feature preservation is proposed, which introduces both feature weighted branch and feature preserving loss. Second, an image degradation model is built based on the blur kernel and noise extracted from the images captured in real scene. Experiments on public datasets show that the proposed SRFCP performs well both in terms of the objective evaluation index and the subjective visual effect. Then, a binocular visual measurement platform is set up to measure the distance of adjacent feature points on a chessboard. Lastly, several SR algorithms are evaluated by the improvement they bring to the measurement accuracy. Experimental results show that the distance measurement performance can be significantly improved by the images reconstructed by the SRFCP. Therefore, the proposed SRFCP can accurately reconstruct the high-resolution images while preserving the features coordinates, which is crucial for the visual measurement in structural monitoring.

Further Exploration of Deep Aggregation for Shadow Detection

Shadow detection is a fundamental challenge in the field of computer vision. It requires the network to understand the global semantics and local details of the image. All existing methods depend on the aggregation of the features of a multi-stage pre-trained convolution neural network but in comparison to high-level capabilities, low-level capabilities provide less to the detection performance. Using low-level features not only increases the complex difficulty of the network but also reduces the time efficiency. In this article, we propose a new shadow detector, which only uses high-level features and explores the complementary information between adjacent feature layers. Experiments show that the technique in this paper can accurately detect shadows and perform well compared with the most advanced methods. The detailed experiments performed in three public shadow detection datasets SUB, UCF, and ISTD demonstrate that the suggested method is efficient and stable.