Enhancing Dense Small Object Detection in UAV Images Based on Hybrid Transformer

Transformer-based models have facilitated significant advances in object detection.However,their extensive computational consumption and suboptimal detection of dense small objects curtail their applicability in unmanned aerial vehicle(UAV)imagery.Addressing these limitations,we propose a hybrid transformer-based detector,H-DETR,and enhance it for dense small objects,leading to an accurate and efficient model.Firstly,we introduce a hybrid transformer encoder,which integrates a convolutional neural network-based cross-scale fusion module with the original encoder to handle multi-scale feature sequences more efficiently.Furthermore,we propose two novel strategies to enhance detection performance without incurring additional inference computation.Query filter is designed to cope with the dense clustering inherent in drone-captured images by counteracting similar queries with a training-aware non-maximum suppression.Adversarial denoising learning is a novel enhancement method inspired by adversarial learning,which improves the detection of numerous small targets by counteracting the effects of artificial spatial and semantic noise.Extensive experiments on the VisDrone and UAVDT datasets substantiate the effectiveness of our approach,achieving a significant improvement in accuracy with a reduction in computational complexity.Our method achieves 31.9%and 21.1%AP on the VisDrone and UAVDT datasets,respectively,and has a faster inference speed,making it a competitive model in UAV image object detection.

Tomato detection method using domain adaptive learning for dense planting environments

This study aimed to address the challenge of accurately and reliably detecting tomatoes in dense planting environments,a critical prerequisite for the automation implementation of robotic harvesting.However,the heavy reliance on extensive manually annotated datasets for training deep learning models still poses significant limitations to their application in real-world agricultural production environments.To overcome these limitations,we employed domain adaptive learning approach combined with the YOLOv5 model to develop a novel tomato detection model called as TDA-YOLO(tomato detection domain adaptation).We designated the normal illumination scenes in dense planting environments as the source domain and utilized various other illumination scenes as the target domain.To construct bridge mechanism between source and target domains,neural preset for color style transfer is introduced to generate a pseudo-dataset,which served to deal with domain discrepancy.Furthermore,this study combines the semi-supervised learning method to enable the model to extract domain-invariant features more fully,and uses knowledge distillation to improve the model's ability to adapt to the target domain.Additionally,for purpose of promoting inference speed and low computational demand,the lightweight FasterNet network was integrated into the YOLOv5's C3 module,creating a modified C3_Faster module.The experimental results demonstrated that the proposed TDA-YOLO model significantly outperformed original YOLOv5s model,achieving a mAP(mean average precision)of 96.80%for tomato detection across diverse scenarios in dense planting environments,increasing by 7.19 percentage points;Compared with the latest YOLOv8 and YOLOv9,it is also 2.17 and 1.19 percentage points higher,respectively.The model's average detection time per image was an impressive 15 milliseconds,with a FLOPs(floating point operations per second)count of 13.8 G.After acceleration processing,the detection accuracy of the TDA-YOLO model on the Jetson Xavier NX development board is 90.95%,the mAP value is 91.35%,and the detection time of each image is 21 ms,which can still meet the requirements of real-time detection of tomatoes in dense planting environment.The experimental results show that the proposed TDA-YOLO model can accurately and quickly detect tomatoes in dense planting environment,and at the same time avoid the use of a large number of annotated data,which provides technical support for the development of automatic harvesting systems for tomatoes and other fruits.

A quantum blind signature scheme based on dense coding for non-entangled states

In some schemes, quantum blind signatures require the use of difficult-to-prepare multiparticle entangled states. By considering the communication overhead, quantum operation complexity, verification efficiency and other relevant factors in practical situations, this article proposes a non-entangled quantum blind signature scheme based on dense encoding. The information owner utilizes dense encoding and hash functions to blind the information while reducing the use of quantum resources. After receiving particles, the signer encrypts the message using a one-way function and performs a Hadamard gate operation on the selected single photon to generate the signature. Then the verifier performs a Hadamard gate inverse operation on the signature and combines it with the encoding rules to restore the message and complete the verification.Compared with some typical quantum blind signature protocols, this protocol has strong blindness in privacy protection,and higher flexibility in scalability and application. The signer can adjust the signature operation according to the actual situation, which greatly simplifies the complexity of the signature. By simultaneously utilizing the secondary distribution and rearrangement of non-entangled quantum states, a non-entangled quantum state representation of three bits of classical information is achieved, reducing the use of a large amount of quantum resources and lowering implementation costs. This improves both signature verification efficiency and communication efficiency while, at the same time, this scheme meets the requirements of unforgeability, non-repudiation, and prevention of information leakage.

Vehicle Abnormal Behavior Detection Based on Dense Block and Soft Thresholding

With the rapid advancement of social economies,intelligent transportation systems are gaining increasing atten-tion.Central to these systems is the detection of abnormal vehicle behavior,which remains a critical challenge due to the complexity of urban roadways and the variability of external conditions.Current research on detecting abnormal traffic behaviors is still nascent,with significant room for improvement in recognition accuracy.To address this,this research has developed a new model for recognizing abnormal traffic behaviors.This model employs the R3D network as its core architecture,incorporating a dense block to facilitate feature reuse.This approach not only enhances performance with fewer parameters and reduced computational demands but also allows for the acquisition of new features while simplifying the overall network structure.Additionally,this research integrates a self-attentive method that dynamically adjusts to the prevailing traffic conditions,optimizing the relevance of features for the task at hand.For temporal analysis,a Bi-LSTM layer is utilized to extract and learn from time-based data nuances.This research conducted a series of comparative experiments using the UCF-Crime dataset,achieving a notable accuracy of 89.30%on our test set.Our results demonstrate that our model not only operates with fewer parameters but also achieves superior recognition accuracy compared to previous models.

Attention-Based Residual Dense Shrinkage Network for ECG Denoising

Electrocardiogram(ECG)signal is one of the noninvasive physiological measurement techniques commonly usedin cardiac diagnosis.However,in real scenarios,the ECGsignal is susceptible to various noise erosion,which affectsthe subsequent pathological analysis.Therefore,the effective removal of the noise from ECG signals has becomea top priority in cardiac diagnostic research.Aiming at the problem of incomplete signal shape retention andlow signal-to-noise ratio(SNR)after denoising,a novel ECG denoising network,named attention-based residualdense shrinkage network(ARDSN),is proposed in this paper.Firstly,the shallow ECG characteristics are extractedby a shallow feature extraction network(SFEN).Then,the residual dense shrinkage attention block(RDSAB)isused for adaptive noise suppression.Finally,feature fusion representation(FFR)is performed on the hierarchicalfeatures extracted by a series of RDSABs to reconstruct the de-noised ECG signal.Experiments on the MIT-BIHarrhythmia database and MIT-BIH noise stress test database indicate that the proposed scheme can effectively resistthe interference of different sources of noise on the ECG signal.

MSADCN:Multi-Scale Attentional Densely Connected Network for Automated Bone Age Assessment

Bone age assessment(BAA)helps doctors determine how a child’s bones grow and develop in clinical medicine.Traditional BAA methods rely on clinician expertise,leading to time-consuming predictions and inaccurate results.Most deep learning-based BAA methods feed the extracted critical points of images into the network by providing additional annotations.This operation is costly and subjective.To address these problems,we propose a multi-scale attentional densely connected network(MSADCN)in this paper.MSADCN constructs a multi-scale dense connectivity mechanism,which can avoid overfitting,obtain the local features effectively and prevent gradient vanishing even in limited training data.First,MSADCN designs multi-scale structures in the densely connected network to extract fine-grained features at different scales.Then,coordinate attention is embedded to focus on critical features and automatically locate the regions of interest(ROI)without additional annotation.In addition,to improve the model’s generalization,transfer learning is applied to train the proposed MSADCN on the public dataset IMDB-WIKI,and the obtained pre-trained weights are loaded onto the Radiological Society of North America(RSNA)dataset.Finally,label distribution learning(LDL)and expectation regression techniques are introduced into our model to exploit the correlation between hand bone images of different ages,which can obtain stable age estimates.Extensive experiments confirm that our model can converge more efficiently and obtain a mean absolute error(MAE)of 4.64 months,outperforming some state-of-the-art BAA methods.

The seismicity in the middle section of the Altyn Tagh Fault system revealed by a dense nodal seismic array

The left-lateral Altyn Tagh Fault(ATF) system is the northern boundary of the Qinghai-Xizang Plateau, separating the Tarim Basin and the Qaidam Basin. The middle section of ATF has not recorded any large earthquakes since1598 AD, so the potential seismic hazard is unclear. We develope an earthquake catalog using continuous waveform data recorded by the Tarim-Altyn-Qaidam dense nodal seismic array from September 17 to November23, 2021 in the middle section of ATF. With the machine learning-based picker, phase association, location, match and locate workflow, we detecte 233 earthquakes with M_L-1–3, far more than 6 earthquakes in the routine catalog. Combining with focal mechanism solutions and the local fault structure, we find that seismic events are clustered along the ATF with strike-slip focal mechanisms and on the southern secondary faults with thrusting focal mechanisms. This overall seismic activity in the middle section of the ATF might be due to the northeastward transpressional motion of the Qinghai-Xizang Plateau block at the western margin of the Qaidam Basin.

Correg-Yolov3:a Method for Dense Buildings Detection in High-resolution Remote Sensing Images

The exploration of building detection plays an important role in urban planning,smart city and military.Aiming at the problem of high overlapping ratio of detection frames for dense building detection in high resolution remote sensing images,we present an effective YOLOv3 framework,corner regression-based YOLOv3(Correg-YOLOv3),to localize dense building accurately.This improved YOLOv3 algorithm establishes a vertex regression mechanism and an additional loss item about building vertex offsets relative to the center point of bounding box.By extending output dimensions,the trained model is able to output the rectangular bounding boxes and the building vertices meanwhile.Finally,we evaluate the performance of the Correg-YOLOv3 on our self-produced data set and provide a comparative analysis qualitatively and quantitatively.The experimental results achieve high performance in precision(96.45%),recall rate(95.75%),F1 score(96.10%)and average precision(98.05%),which were 2.73%,5.4%,4.1%and 4.73%higher than that of YOLOv3.Therefore,our proposed algorithm effectively tackles the problem of dense building detection in high resolution images.

Shear wave splitting analysis of local earthquakes from dense arrays in Shimian,Sichuan

The Shimian area of Sichuan sits at the junction of the Bayan Har block.Sichuan-Yunnan rhombic block,and Yangtze block,where several faults intersect.This region features intense tectonic activity and frequent earthquakes.In this study,we used local seismic waveform data recorded using dense arrays deployed in the Shimian area to obtain the shear wave splitting parameters at 55 seismic stations and thereby determine the crustal anisotropic characteristics of the region.We then analyzed the crustal stress pattern and tectonic setting and explored their relationship in the study area.Although some stations returned a polarization direction of NNW-SSE.a dominant polarization direction of NW-SE was obtained for the fast shear wave at most seismic stations in the study area.The polarization directions of the fast shear wave were highly consistent throughout the study-area.This orientation was in accordance with the direction of the regional principal compressive stress and parallel to the trend of the Xianshuihe and Daliangshan faults.The distribution of crustal anisotropy in this area was affected by the regional tectonic stress field and the fault structures.The mean delay time between fast and slow shear waves was 3.83 ms/km.slightly greater than the values obtained in other regions of Sichuan.This indicates that the crustal media in our study area had a high anisotropic strength and also reveals the influence of tectonic complexity resulting from the intersection of multiple faults on the strength of seismic anisotropy.

Siamese Dense Pixel-Level Fusion Network for Real-Time UAV Tracking

Onboard visual object tracking in unmanned aerial vehicles(UAVs)has attractedmuch interest due to its versatility.Meanwhile,due to high precision,Siamese networks are becoming hot spots in visual object tracking.However,most Siamese trackers fail to balance the tracking accuracy and time within onboard limited computational resources of UAVs.To meet the tracking precision and real-time requirements,this paper proposes a Siamese dense pixel-level network for UAV object tracking named SiamDPL.Specifically,the Siamese network extracts features of the search region and the template region through a parameter-shared backbone network,then performs correlationmatching to obtain the candidate regionwith high similarity.To improve the matching effect of template and search features,this paper designs a dense pixel-level feature fusion module to enhance the matching ability by pixel-wise correlation and enrich the feature diversity by dense connection.An attention module composed of self-attention and channel attention is introduced to learn global context information and selectively emphasize the target feature region in the spatial and channel dimensions.In addition,a target localization module is designed to improve target location accuracy.Compared with other advanced trackers,experiments on two public benchmarks,which are UAV123@10fps and UAV20L fromthe unmanned air vehicle123(UAV123)dataset,show that SiamDPL can achieve superior performance and low complexity with a running speed of 100.1 fps on NVIDIA TITAN RTX.

Low Complexity Joint Spectrum Resource and Power Allocation for Ultra Dense Networks

In this paper,we propose a low complexity spectrum resource allocation scheme cross the access points(APs)for the ultra dense networks(UDNs),in which all the APs are divided into several AP groups(APGs)and the total bandwidth is divided into several narrow band spectrum resources and each spectrum resource is allocated to APGs independently to decrease the interference among the cells.Furthermore,we investigate the joint spectrum and power allocation problem in UDNs to maximize the overall throughput.The problem is formulated as a mixed-integer nonconvex optimization(MINCP)problem which is difficult to solve in general.The joint optimization problem is decomposed into two subproblems in terms of the spectrum allocation and power allocation respectively.For the spectrum allocation,we model it as a auction problem and a combinatorial auction approach is proposed to tackle it.In addition,the DC programming method is adopted to optimize the power allocation subproblem.To decrease the signaling and computational overhead,we propose a distributed algorithm based on the Lagrangian dual method.Simulation results illustrate that the proposed algorithm can effectively improve the system throughput.

Spectrum-Efficient User Association in Caching Enabled Dense Small Cell Network

In this paper,we reveal the fundamental limitation of network densification on the performance of caching enabled small cell network(CSCN)under two typical user association rules,namely,contentand distance-based rules.It indicates that immoderately caching content would significantly change the interference distribution in CSCN,which may degrade the network area spectral efficiency(ASE).Meanwhile,it is shown that content-based rule outperforms the distance-based rule in terms of network ASE only when small cell base stations(BSs)are sparsely deployed with low decoding thresholds.Moreover,it is proved that network ASE under distance-based user association serves as the upper bound of that under content-based rule in dense BS regime.To enable more spectrum-efficient user association in dense CSCN,we further optimize network ASE by designing a probabilistic content retrieving strategy based on distance-based rule.With the optimized retrieving probability,network ASE could be substantially enhanced and even increase with the growing BS density in dense BS regime.

Dense Spatial-Temporal Graph Convolutional Network Based on Lightweight OpenPose for Detecting Falls

Fall behavior is closely related to high mortality in the elderly,so fall detection becomes an important and urgent research area.However,the existing fall detection methods are difficult to be applied in daily life due to a large amount of calculation and poor detection accuracy.To solve the above problems,this paper proposes a dense spatial-temporal graph convolutional network based on lightweight OpenPose.Lightweight OpenPose uses MobileNet as a feature extraction network,and the prediction layer uses bottleneck-asymmetric structure,thus reducing the amount of the network.The bottleneck-asymmetrical structure compresses the number of input channels of feature maps by 1×1 convolution and replaces the 7×7 convolution structure with the asymmetric structure of 1×7 convolution,7×1 convolution,and 7×7 convolution in parallel.The spatial-temporal graph convolutional network divides the multi-layer convolution into dense blocks,and the convolutional layers in each dense block are connected,thus improving the feature transitivity,enhancing the network’s ability to extract features,thus improving the detection accuracy.Two representative datasets,Multiple Cameras Fall dataset(MCF),and Nanyang Technological University Red Green Blue+Depth Action Recognition dataset(NTU RGB+D),are selected for our experiments,among which NTU RGB+D has two evaluation benchmarks.The results show that the proposed model is superior to the current fall detection models.The accuracy of this network on the MCF dataset is 96.3%,and the accuracies on the two evaluation benchmarks of the NTU RGB+D dataset are 85.6%and 93.5%,respectively.

Novel Path Counting-Based Method for Fractal Dimension Estimation of the Ultra-Dense Networks

Next-generation networks,including the Internet of Things(IoT),fifth-generation cellular systems(5G),and sixth-generation cellular systems(6G),suf-fer from the dramatic increase of the number of deployed devices.This puts high constraints and challenges on the design of such networks.Structural changing of the network is one of such challenges that affect the network performance,includ-ing the required quality of service(QoS).The fractal dimension(FD)is consid-ered one of the main indicators used to represent the structure of the communication network.To this end,this work analyzes the FD of the network and its use for telecommunication networks investigation and planning.The clus-ter growing method for assessing the FD is introduced and analyzed.The article proposes a novel method for estimating the FD of a communication network,based on assessing the network’s connectivity,by searching for the shortest routes.Unlike the cluster growing method,the proposed method does not require multiple iterations,which reduces the number of calculations,and increases the stability of the results obtained.Thus,the proposed method requires less compu-tational cost than the cluster growing method and achieves higher stability.The method is quite simple to implement and can be used in the tasks of research and planning of modern and promising communication networks.The developed method is evaluated for two different network structures and compared with the cluster growing method.Results validate the developed method.

Speech Enhancement via Mask-Mapping Based Residual Dense Network

Masking-based and spectrum mapping-based methods are the two main algorithms of speech enhancement with deep neural network(DNN).But the mapping-based methods only utilizes the phase of noisy speech,which limits the upper bound of speech enhancement performance.Maskingbased methods need to accurately estimate the masking which is still the key problem.Combining the advantages of above two types of methods,this paper proposes the speech enhancement algorithm MM-RDN(maskingmapping residual dense network)based on masking-mapping(MM)and residual dense network(RDN).Using the logarithmic power spectrogram(LPS)of consecutive frames,MM estimates the ideal ratio masking(IRM)matrix of consecutive frames.RDN can make full use of feature maps of all layers.Meanwhile,using the global residual learning to combine the shallow features and deep features,RDN obtains the global dense features from the LPS,thereby improves estimated accuracy of the IRM matrix.Simulations show that the proposed method achieves attractive speech enhancement performance in various acoustic environments.Specifically,in the untrained acoustic test with limited priors,e.g.,unmatched signal-to-noise ratio(SNR)and unmatched noise category,MM-RDN can still outperform the existing convolutional recurrent network(CRN)method in themeasures of perceptual evaluation of speech quality(PESQ)and other evaluation indexes.It indicates that the proposed algorithm is more generalized in untrained conditions.

3D S-wave velocity structure of the Ningdu basin in Jiangxi province inferred from ambient noise tomography with dense array

The Ningdu basin,located in southern Jiangxi province of southwest China,is one of the Mesozoic basin groups which has exploration prospects for geothermal energy.A study on the detailed velocity structure of the Ningdu basin can provide important information for geothermal resource exploration.In this study,we deployed a dense seismic array in the Ningdu basin to investigate the 3D velocity structure and discuss implications for geothermal exploration and geological evolution.Based on the dense seismic array including 35 short-period(5 s-100 Hz)seismometers with an average interstation distance of~5 km,Rayleigh surface wave dispersion curves were extracted from the continuous ambient noise data for surface wave tomographic inversion.Group velocity tomography was conducted and the 3D S-wave velocity structure was inverted by the neighborhood algorithm.The results revealed obvious low-velocity anomalies in the center of the basin,consistent with the low-velocity Cretaceous sedimentary rocks.The basement and basin-controlling fault can also be depicted by the S-wave velocity anomalies.The obvious seismic interface is about 2 km depth in the basin center and decreases to 700 m depth near the basin boundary,suggesting spatial thickness variations of the Cretaceous sediment.The fault features of the S-wave velocity profile coincide with the geological cognition of the western boundary basincontrolling fault,which may provide possible upwelling channels for geothermal fluid.This study suggests that seismic tomography with a dense array is an effective method and can play an important role in the detailed investigations of sedimentary basins.

Possible Nature of an Electron Dense Substance in the Thylakoid Lumen of Chloroplasts

The fixation of leaves of Tanacetum vulgare L. in glutaraldehyde makes it possible to isolate chloroplasts without sacrificing an electron dense substance in the thylakoid lumen. The extraction of lipids from the chloroplasts isolated from the leaves preliminarily fixed in glutaraldehyde has demonstrated that glycerolipids (galactolipids and phospholipids) are not manifested in TLC, whereas isoprenoid lipids (chlorophyll, carotenoids) are manifested. Presumably, isoprenoid lipids are not fixed with glutaraldehyde and are extracted from the thylakoid membrane. The ultrastructural control demonstrates that the electron dense substance from the thylakoid lumen is also extracted. It is possible that this substance is of isoprenoid nature.

Design of dense gap-graded friction course mixture

The design procedure of a dense gap-graded friction course（DGGFC） with coarse aggregate void filling method is presented. Testing results show that a DGGFC mixture possesses a dense stone-matrix structure, good stability and almost the same texture depth as stone matrix asphalt （SMA）. It also has a coarse and even surface after paving and has no separation during construction. It is durable and impermeable. It balances and improves the inherent inconsistency of asphalt mixture between the large texture depth for skid resistance and the impermeability for durability. The actual application in the Nanning-Liuzhou Expressway also shows that the performance of the DGGFC is as excellent as that of SMA, while the DGGFC mixture is cheaper than SMA. The DGGFC mixture is good for wearing course of pavement. Further research on DGGFC can be helpful for improving the surface skid resistance, prolonging the life-span period and reducing the construction costs of asphalt pavement.

Analysis on the Cause of Dense Fog in Jining on November 10,2005

Through analyzing the situation of atmospheric circulation and the variation of meteorological element at the time of dense fog happened on November 10,2005 in Jining,the results indicated that there were abundant water vapor,weaker pressure field and wind field,weaker cold air in high altitude,stable stratification in underlying bed and ground was located in the bottom of cold high pressure,all those conditions were beneficial to the generation of dense fog in Jining.

Extracting and evaluating method of web dense cores

This paper focuses on some key problems in web community discovery and link analysis.Based on the topic-oriented technology,the characteristics of a bipartite graph are studied.An Х bipartite core set is introduced to more clearly define extracting ways.By scanning the topic subgraph to construct Х bipartite graph and then prune the graph with i and j ,an Х bipartite core set,which is also the minimum element of a community,can be found.Finally,a hierarchical clustering algorithm is applied to many Х bipartite core sets and the dendrogram of the community inner construction is obtained.The correctness of the constructing and pruning method is proved and the algorithm is designed.The typical datasets in the experiment are prepared according to the way in HITS（hyperlink-induced topic search）.Ten topics and four search engines are chosen and the returned results are integrated.The modularity,which is a measure of the strength of the community structure in the social network,is used to validate the efficiency of the proposed method.The experimental results show that the proposed algorithm is effective and efficient.