Channel Characteristics Research Articles

AeroNet: Efficient YOLOv7 for Tiny-Object Detection in UAV Imagery

The detection of multiple tiny objects from diverse perspectives using unmanned aerial vehicles (UAVs) and onboard edge devices presents a significant challenge in computer vision. To address that, this study proposes AeroNet, a lightweight and efficient detection algorithm based on YOLOv7 (You Only Look Once version7).This algorithm features the LHGNet (Lightweight High-Performance GhostNet) backbone, an advanced feature extraction network that integrates depth-wise separable convolution and channel shuffle modules.These modules enable deeper exploration of network features, promoting the fusion of local detail information and channel characteristics. Additionally, this research introduces the LGS(Lightweight Gradient-Sensitive) bottleneck and LGSCSP(Lightweight Gradient-Sensitive Cross Stage Partial Network) fusion module in the neck to reduce computational complexity while maintaining accuracy. Structural modifications and adjusted feature map sizes further enhance detection accuracy. Evaluated on the SkyFusion dataset,this method demonstrated a 25.0% reduction in parameter count and a 12.8% increase in mAP (0.5) compared to YOLOv7. These results underscore the effectiveness of this proposed approach in improving detection accuracy and model efficiency through the proposed enhancements.

Based on the N2N-SAMP for sparse underwater acoustic channel estimation

IntroductionOrthogonal Frequency Division Multiplexing (OFDM) is widely recognized for its high efficiency in modulation techniques and has been extensively applied in underwater acoustic communication. However, the unique sparsity and noise interference characteristics of the underwater channel pose significant challenges to the performance of traditional channel estimation methods.MethodsTo address these challenges, we propose a sparse underwater channel estimation method that combines the Noise2Noise (N2N) algorithm with the Sparsity Adaptive Matching Pursuit (SAMP) algorithm. This novel approach integrates the N2N technique from image denoising theory with the SAMP algorithm, utilizing a constant iteration termination threshold that does not require prior information. The method leverages the U-net neural network structure to denoise noisy pilot signals, thereby restoring channel sparsity and enhancing the accuracy of channel estimation.ResultsSimulation results indicate that our proposed method demonstrates commendable channel estimation performance across various signal-to-noise ratio (SNR) conditions. Notably, in low SNR environments, the N2N-SAMP algorithm significantly outperforms the traditional SAMP algorithm in terms of Mean Squared Error (MSE) and Bit Error Rate (BER). Specifically, at SNR levels of 0 dB, 10 dB, and 20 dB, the MSE of channel estimation is reduced by 58.95%, 76.08%, and 19.42%, respectively, compared to the SAMP algorithm that selects the optimal threshold based on noise power. Furthermore, the system’s BER is decreased by 12.35%, 26.41%, and 29.62%, respectively.DiscussionThe findings suggest that the integration of N2N and SAMP algorithms offers a promising solution for improving channel estimation in underwater communication channels, especially under low SNR conditions. The significant reduction in MSE and BER highlights the effectiveness of our proposed method in enhancing the reliability and accuracy of underwater communication systems.

Features of Motion and Heat Transfer of Swirling Flows in Channels of Complex Geometry

The computational and experimental study results of swirling single-phase coolant motion and heat transfer for the standard operation parameters of a nuclear power plant are presented. The experimental model is a vertical heat exchanger of a “pipe in a pipe” type with the countercurrent movement of coolants. Six different swirlers (three with a constant twist pitch and three with a variable pitch) were considered. The heat exchanger temperature field was measured at various combinations of coolant flow rates, and a channel pressure drop for each swirl was determined. Computational studies were performed using the Omega-based Reynolds stress model and SST model with a correction for curvature streamlines. A good agreement between numerical and experimental data was obtained. Based on the velocity and temperature fields, swirling flow motion features in channels with a variable swirl pitch were discovered. For each intensifier, the effectiveness criterion in comparison with a pipe channel was determined.

Mapping Lithology with Hybrid Attention Mechanism–Long Short-Term Memory: A Hybrid Neural Network Approach Using Remote Sensing and Geophysical Data

Remote sensing provides an efficient roadmap in geological analysis and interpretation. However, some challenges arise when remote sensing techniques are integrated with machine learning in geological surveys. Factors including irregular spatial distribution, sample imbalance, interclass resemblances, regolith, and geochemical similarities impede geological feature diagnosis, interpretation, and identification across varied remote sensing datasets. To address these limitations, a hybrid-attention-integrated long short-term memory (LSTM) network is employed to diagnose, interpret, and identify lithological feature representations in a remote sensing-based geological analysis using multisource data fusion. The experimental design integrates varied datasets including Sentinel-2A, Landsat-9, ASTER, ALOS PALSAR DEM, and Bouguer anomaly gravity data. The proposed model incorporates a hybrid attention mechanism (HAM) comprising channel and spatial attention submodules. HAM utilizes an adaptive technique that merges global-average-pooled features with max-pooled features, enhancing the model’s accuracy in identifying lithological units. Additionally, a channel separation operation is employed to allot refined channel features into clusters based on channel attention maps along the channel dimension. The comprehensive analysis of results from comparative extensive experiments demonstrates HAM-LSTM’s state-of-the-art performance, outperforming existing attention modules and attention-based models (ViT, SE-LSTM, and CBAM-LSTM). Comparing HAM-LSTM to baseline LSTM, the HAM module’s integrated configurations equip the proposed model to better diagnose and identify lithological units, thereby increasing the accuracy by 3.69%.

Quantitative Characterization of Channel Morphology and Main Controlling Factor of Shallow Water Delta—A Case from Ganjiang Delta, Jiangxi, China

(1) This paper selects the modern delta formed by the Ganjiang tributary in Poyang Lake. By performing high density statistical analysis of distribution channel parameters in the area using satellite images and geographic information processing software (LucaSpaceViewer 4.5.2, ArcGIS Pro 3.0.2, Global Mapper v23.1), including length, width, bifurcation angle, bifurcation frequency, and channel sinuosity, the distribution characteristics of delta distribution channels are derived and quantitatively characterized. (2) Classification and evaluation of these characteristics are carried out using factor and cluster analysis, ultimately identifying controlling factors affecting the morphology and distribution of the distribution channels. By statistically analyzing the geometric and bifurcation data of the channels, factor and cluster analysis for data reduction and classification, the channel is finally divided into three categories: Type I channels have relatively high channel length, width, sinuosity, bending amplitude, and a lower bifurcation (or confluence) growth rate; Type II channels are characterized by low channel length, moderate channel width, low sinuosity, low bending amplitude, and a high bifurcation (or confluence) growth rate; Type III channels are defined by moderate channel length, low width, high sinuosity, high bending amplitude, and low bifurcation (or confluence) frequency. (3) After excluding the influence of other factors, it was found that the main controlling factor for the morphology of the Ganjiang Delta channel is flow velocity, which is influenced by changes in the terrain slope. Flow velocity directly affects channel sinuosity, bending amplitude, and bifurcation (or confluence) frequency, and indirectly affects channel length and width.

IDENTIFIKASI KONDISI BANJIR DAN KARAKTERISTIK SALURAN DRAINASE DI KAWASAN KECAMATAN DENPASAR BARAT

West Denpasar, as part of Denpasar City, often faces flood problems caused by weather factors, buildings, and inadequate drainage infrastructure. This article aims to identify flood conditions and assess the characteristics of drainage channels in the areas of Jalan Gunung Tangkuban Perahu, Jalan Gunung Salak, and Jalan Pura Demak. This research was carried out through field observation, measurement, and data analysis regarding rainfall and drainage channel characteristics. The results of qualitative data show that 80% of the local community has experienced flooding, with the majority complaining of clogged drainage conditions. Meanwhile, quantitative data shows that the drainage channel on Jalan Pura Demak has an average width of only 25 cm, which worsens flood conditions with inundation depths reaching 70 cm for more than 6 hours after rain. The conclusion of this study emphasizes the need to increase the maintenance and improvement of drainage infrastructure, as well as community involvement in maintaining the cleanliness of channels to reduce the risk of repeated floods.

Non-target feature filtering for weakly supervised semantic segmentation

Weakly supervised semantic segmentation (WSSS) utilizes weak labels to learn semantic segmentation models, significantly reducing reliance on pixel-level annotations. WSSS typically employs a multi-label classification network to extract image features for constructing localization maps. The quality of the localization map critically influences the performance of WSSS. However, non-target semantic noise within the features impedes the improvement of localization map quality. To address this issue, we propose a non-target feature filtering class activation mapping (NFF-CAM) for WSSS, which can reduce non-target semantic signals and generate higher-quality localization maps. Specifically, the class-constrained dual cosine clustering (CDCC) and channel identification (CI) modules are introduced in NFF-CAM. CDCC effectively addresses the issue of unsuitability in the clustering group relationships of the original features under specified class conditions. CI can efficiently identify channel features containing non-target semantic information. We conduct extensive evaluations of NFF-CAM on popular datasets, including PASCAL VOC 2012 and MS COCO 2014. Experimental results show that NFF-CAM can effectively improve the segmentation performance of off-the-shelf methods.

Base-level and Litho-structural Controls on the Incised Sinuous Bedrock Rivers of the Jawhar Plateau, North Konkan, India

ABSTRACT The coastal rivers draining the mid-altitude Jawhar Plateau in north Konkan display some unusual drainage and channel characteristics. These rivers, especially in their middle reach, are highly sinuous and deeply incised in bedrock basalt. In this study, various geomorphic indices were calculated and interpreted for five such bedrock meandering rivers, namely Purna, Ambika, Auranga, Par and Damanganga. Morphometric data reveal that in spite of more drainage basin area in the mid-elevation category, the rivers are not only deeply entrenched within bedrock basalt, but also have unusually high sinuosity. There are indications of subtle to significant control of dykes and lineaments on the channel planform morphology and channel bed profile. However, various other geomorphic evidences suggest that an increase in the regional slope, induced by a fall in the erosional base-level, has most likely played an important role on longer timescale in the initiation and formation of the incised bedrock meanders in this part of the Deccan Traps region.

Multifeature extraction based MobileViTv3 model for fish feeding behavior recognition from video streaming

The recognition of fish feeding behavior based on machine vision is essential for optimizing fish feeding strategies and enhancing the efficiency of aquaculture. Building an efficient, multi-feature extraction model for fish feeding recognition, especially on mobile and edge devices, remains a significant challenge. In the paper, we propose a novel multi-feature extraction (MFE)-MobileViTv3 model, which improve MobileViTv3 with the MFE blocks. It can extract spatio-temporal features while obtaining the lightweight characteristic. The MFE block is designed by improving ActionNet, with the frequency channel attention (FCA) and multi-head self-attention (MHSA) mechanisms. It can fully extract spatio-temporal, motion, and channel features from video streaming, thereby further improving the feature extraction capabilities, which subsequently enhances the model's recognition accuracy. The experiments were carried out on an industrial aquaculture farm. We built a dataset of Micropterus salmoides, and then conducted the compare experiments. Compared with C3D, R3D, ResNet50, SlowFast, AlexNet, and MobileNetV3, our model can achieve a classification accuracy of 96.7 % for feeding intensity, with fewer parameters (0.96 M) and FLOPs (8.44G). The results show that the proposed model can effectively recognize fish feeding behavior with fewer parameters. Additionally, we introduce two evaluation metrics for the feeding process: Average Feeding Intensity and Strong Feeding Ratio. The metrics are conducive to the quantitative evaluation of the fish's vigor and health status.

Multifrequency Wireless Channel Measurements and Characterization in Indoor Industrial Scenario

Multifrequency Wireless Channel Measurements and Characterization in Indoor Industrial Scenario

Driver Cognitive Distraction Detection based on eye movement behavior and integration of multi-view space-channel feature

The significance of Driver Distraction Detection (DDD) in enhancing road safety is immense and undeniable. While numerous DDD methods concentrate on conventional distractions, such as driving behavior, gaze direction, or hand movements, they often overlook cognitive distractions owing to their intricate nature. This study aims to bridge this gap by leveraging driver eye movement data to detect cognitive distractions. To achieve this, we introduce the Driver Cognitive Distraction Detection (DCDD) model. This model incorporates the driver’s gaze points, captured by an eye tracker, and the DashCam Image (DCI) as inputs. To capture the temporal dynamics of the driver’s eye movements, we devise a preprocessing technique that generates an eye movement heat Map (Map), which highlights the driver’s areas of focus within the DCI. Then, a Fusion Adversarial Network (FAN) is designed by utilizing the Sigmoid activation function to blend DCI features with those of the Map. Subsequently, a shared parameter network is employed for adversarial learning, focusing on the inversely activated features. To further achieve the extraction of DCI and Map features, we propose the Multi-View Space Channel Network (MSCN). This network can integrate low-dimensional spatial features and high-dimensional channel features from different views of DCI and Map to help Transformer capture DCI and Map features more comprehensively in the temporal dimension. By effectively integrating DCI and Map data and recursively extracting time information, our DCDD model achieves remarkable accuracy while controlling model Params and FLOPs, attaining a high score of 96.42%.

An improved multi-scale feature extraction network for medical image segmentation.

The use of U-Net and its variations has led to significant advancements in medical image segmentation. However, the encoder-decoder structures of these models often lose spatial information during downsampling. Skip connections can help address this issue; however, they may also introduce excessive irrelevant background information. Additionally, medical images display significant scale variations and complex tissue structures, making it challenging for existing models to accurately separate tissues from the background. To address these issues, we developed the Res2Net-ConvFormer-Dilation-UNet (Res2-CD-UNet), a multi-scale feature extraction network for medical image segmentation. This study presents a novel U-shaped segmentation network that employs Res2Net as the backbone and incorporates a convolution-style transformer in the encoding stage to enhance global attention. Additionally, a novel channel feature fusion block (CFFB) has been introduced in the skip connection stage to minimize the effects of background noise. The proposed model was evaluated using publicly available datasets, Synapse and Seg.A.2023. Using the Synapse dataset, the average dice similarity coefficient (DSC) reached 83.92%, which was 1.96% higher than the suboptimal model, and the average Hausdorff distance (HD) was 14.51 mm. Among the eight organs evaluated, optimal results were achieved for four organs. Similarly, using the Seg.A.2023 dataset, the proposed model also achieved the best results with an average DSC of 93.27%. The results of this study indicate that the proposed model can more accurately segment regions of interest and better extract multi-scale features in medical images than existing deep-learning algorithms.

Manifestation of Luttinger liquid effects in a hybrid metal-semiconductor double-quantum dot device

We theoretically study the transport properties of a hybrid nanodevice comprised of two large metallic islands incorporated in a two-dimensional electron gas. The high-tunability of the conducting channels electrically connecting two islands to each other and to the leads allows us to treat the setup as a realization of a multi-channel two-site charge Kondo (2SCK) model. It is shown that the leading temperature dependence of the conductance in the 2SCK circuit satisfies the conductance scaling of a single-impurity problem in a Luttinger liquid, whose interaction parameter is fully determined by the number of conducting channels in the device. We demonstrate that the finite weak backscattering in all conducting channels features the appearance of the sub-leading temperature dependencies in linear conductance. At the special critical point, we predict an equivalency between the 2SCK nanodevice and a single-site two-channel charge Kondo problem, where one Kondo channel is implemented by a non-interacting electron gas and the second Kondo channel is attributed to the Luttinger liquid.

The Influence of Non-Pressure Pipes Filled to Different Degrees on the Hydraulic Characteristics of Channel–Pipe Combined Irrigation Systems

Modern irrigation areas often use a combination of channels and pipes for irrigation. Due to changes in terrain, inflow, and pipe diameter, pipelines are often prone to experiencing a state of no pressure. This lack of pressure affects not only the velocity distribution and water surface profile of the main channel but also the velocity distribution and flow distribution of the pipeline. Therefore, in this study, we employed a combination of physical model experiments and theoretical analysis to study the influence of non-pressure pipelines on the hydraulic characteristics of channel–pipe combined irrigation systems filled to different degrees. Through this, the variation laws of the flow velocity distribution, turbulence intensity, water surface, diversion width, and diversion ratio under different filling degrees were obtained. The non-pressure pipeline flow velocity expression was obtained through dimensional analysis, and the calculation formula for the non-pressure pipeline flow velocity coefficient was fitted using linear regression analysis. The relative error between the calculated value and the measured value did not exceed 8.81%. The research results presented in this article can provide technical support for the design and maintenance of channel–pipe combined irrigation systems.

A comprehensive investigation of PRMT5 in the prognosis and ion channel features of lung cancer

A comprehensive investigation of PRMT5 in the prognosis and ion channel features of lung cancer

Microwave Irradiation-Assisted Synthesis of Anisotropic Crown Ether-Grafted Bamboo Pulp Aerogel as a Chelating Agent for Selective Adsorption of Heavy Metals (Mn+)

Crown ether is widely used in water purification because of its ring structure and good selective adsorption of specific heavy metals. However, its high cost and difficulty in recycling limit the purification of heavy metals in water. The anisotropic [2,4]-dibenzo-18-crown-6-modified bamboo pulp aerogel (DB18C6/PA) is successfully synthesized by microwave irradiation and directional freezing technology. The physical and chemical properties of DB18C6/PA are analyzed by FTIR, XPS, SEM, TEM, TGA, surface area and porosity analyzers. Single or multivariate systems containing Pb2+, Cu2+ and Cd2+ are used as adsorbents. The effects of the DB18C6 addition amount, pH, initial concentration and adsorption temperature on the adsorption of DB18C6/PA are systematically explored. Pseudo-first-order kinetic models, pseudo-second-order kinetic models and the isothermal adsorption models of Langmuir and Freundlich are used to fit the experimental data. The adsorption selectivity is analyzed from the distribution coefficient and the separation factor, and the adsorption mechanism is discussed. The results show that anisotropic DB18C6/PA has the characteristics of 3D directional channels, high porosity (97.67%), large specific surface area (103.7 m2/g), good thermal stability and regeneration (the number of cycles is greater than 5). The surface has a variety of functional groups, including a hydroxyl group, aldehyde group, ether bond, etc. In the single and multivariate systems of Pb2+, Cu2+ and Cd2+, the adsorption process of DB18C6/PA conforms to the pseudo-second-order kinetic model, and the results conform to the Freundlich adsorption isothermal model (a few of them conformed to the Langmuir adsorption isothermal model), indicating that chemical adsorption and physical adsorption are involved in the adsorption process, and the adsorption process is a spontaneous endothermic process. In the single solution system, the maximum adsorption capacities of Pb2+, Cu2+ and Cd2+ by DB18C6/PA are 129.15, 29.85 and 27.89 mg/g, respectively. The adsorption selectivity of DB18C6/PA on Pb2+, Cu2+ and Cd2+ is in the order of Pb2+ >> Cu2+ > Cd2+.

Edge-guided and hierarchical aggregation network for robust medical image segmentation

Medical image segmentation with the convolutional neural networks (CNNs), significantly enhances clinical analysis and disease diagnosis. However, medical images inherently exhibit large intra-class variability, minimal inter-class differences, and substantial noise levels. Extracting robust contextual information and aggregating discriminative features for fine-grained segmentation remains a formidable task. Additionally, existing methods often struggle with producing accurate mask edges, leading to blurred boundaries and reduced segmentation precision. This paper introduces a novel Edge-guided and Hierarchical Aggregation Network (EHANet) which excels at capturing rich contextual information and preserving fine spatial details, addressing the critical issues of inaccurate mask edges and detail loss prevalent in current segmentation models. The Inter-layer Edge-aware Module (IEM) enhances edge prediction accuracy by fusing early encoder layers, ensuring precise edge delineation. The Efficient Fusion Attention Module (EFA) adaptively emphasizes critical spatial and channel features while filtering out redundancies, enhancing the model’s perception and representation capabilities. The Adaptive Hierarchical Feature Aggregation Module (AHFA) module optimizes feature fusion within the decoder, maintaining essential information and improving reconstruction fidelity through hierarchical processing. Quantitative and qualitative experiments on four public datasets demonstrate the effectiveness of EHANet in achieving superior mIoU, mDice, and edge accuracy against eight other state-of-the-art segmentation methods, highlighting its robustness and precision in diverse clinical scenarios.

Pretending to be a VIP! Characterization and Detection of Fake and Clone Channels on Telegram

Telegram is a widely used instant messaging app that has gained popularity due to its high level of privacy protection. Telegram has standout social network features like channels, which are virtual rooms where only administrators can post and broadcast messages to all subscribers. However, these same features have also led to the emergence of problematic activities and a significant number of fake accounts. To address these issues, Telegram has introduced verified and scam marks for channels, but only a small number of official channels are currently marked as verified, and only a few fakes as scams. In this research, we conduct a large-scale analysis of Telegram by collecting data from 120,979 different public channels and over 247 million messages. We identify and analyze two types of channels: Clones and fakes. Clones are channels that publish identical content from another channel in order to gain subscribers and promote services. Fakes, on the other hand, are channels that impersonate celebrities or well-known services by posting their own messages. To automatically detect fake channels, we propose a machine learning model that achieves an F1-score of 85.45%. By applying this model to our dataset, we find the main targets of fakes are political figures, well-known people such as actors or singers, and services.

Cross-layer Authentication Mechanism Model Combined with 5G Converged Channel Fingerprint

In the actual communication environment, once the attacker successfully steals the legitimate channel information, the key information can be cracked according to the authentication response, so there is a security risk of key leakage. This paper combines the physical layer channel characteristics with the shared key, and combines it into a joint key to design a challenge-response physical layer authentication mechanism based on interpolation polynomial. Then, this method is applied to the EAP-AKA’ authentication protocol of 5G network, and a cross-layer authentication mechanism for 5G converged channel fingerprint is proposed. Finally, using the captured high-level authentication challenge response data, a simulation environment is built in MATLAB and the feasibility and security of the scheme are verified. The experimental results show that the authentication mechanism has better authentication performance and security performance. Compared with the traditional high-level authentication mechanism, it has certain advantages in computing overhead and can meet the 5G application scenarios of large-scale IoT terminals. Moreover, it combines physical layer authentication with high-level authentication, realizes mutual complementarity and mutual enhancement, and enhances the security performance of authentication.

AFCF-Net: A novel U-Net based asymmetric feature calibration and fusion network for skin lesion image segmentation.

Skin lesion segmentation plays a pivotal role in the diagnosis and treatment of skin diseases. By using deep neural networks to segment lesion areas, doctors can more accurately assess the severity of health-related conditions of patients and promptly implement appropriate treatment measures, thereby enhancing treatment outcomes and improving the quality of life (QoL) of patients. However, existing segmentation networks still face challenges in balancing segmentation performance and efficiency. To address this issue, a novel network, named AFCF-Net, is proposed in this paper for skin lesion segmentation tasks. Firstly, the proposed network employs a newly designed spatial channel feature calibration convolution (SCFCConv) to enhance its ability to perceive spatial and channel features. Secondly, AFCF-Net utilizes newly designed feature symmetric fusion convolution (FSFConv) in skip connections to selectively fuse features from different levels, thereby enhancing its sensitivity to texture, edges, and other detailed features. In addition, a feature attention recombination module (FARM) is added to the bottleneck of the proposed network to comprehensively acquire and utilize contextual information at different scales, thus improving the network's generalization ability. Finally, a newly designed multi-level feature aggregation branch is introduced as an additional decoder for AFCF-Net to supplement key features lost during the original decoding process. Experiments, conducted on four skin image datasets, demonstrate that the proposed AFCF-Net network achieves better segmentation performance with fewer parameters and computational resources, compared to state-of-the-art segmentation networks. Additionally, AFCF-Net exhibits stronger generalization ability.