Search

The early diagnosis of any disease can be curable with a small amount of human effort. Most people will not be able to detect their illness before it becomes chronic. Breast cancer has been one of the diseases that can be treated until it spreads to all parts of the body when the disease is diagnosed at an earlier stage.Breast cancer is causing an increase in death rates all over the world. The medical practitioner can diagnose diseases incorrectly due to misinterpretation. In both developed and emerging countries, breast cancer is one of the leading causes of death for women. Breast cancer can be detected and classified early on in its progression, allowing patients to get adequate treatment. For detecting and classifying breast cancer in breast cytology videos, deep learning frameworks are suggested. The suggested architecture uses the Convolutional layer and the pooling layer to remove features. The data would be classified using the dense layer.The Convolutional Neural Network and DenseNet are used to extract image features that are then fed into a fully connected layer for the classification of malignant and benign cellsFor image classification of breast cancer, the models are trained, validated, and tested.The distinction is based on CNN and DenseNet's different batch sizes and learning rates. To see how well the proposed CNN and DenseNet programs work, they are compared. In terms of accuracy, the suggested system will outperform CNN and DenseNet in the diagnosis and classification of breast tumors from histological photographs. Hyper parameter tuning is used to measure the efficiency of CNN and DenseNet.

Due to the excessive concentration of computing resources in the traditional centralized cloud service system, there will be three prominent problems of management confusion, construction cost and network delay. Therefore, we propose to virtualize regional edge computing resources in intelligent buildings as edge service pooling, then presents a hierarchical cloud platform with dual-service pooling structure and a dynamic strategy for the proposed model. The analytic hierarchy process (AHP) based quality of service (QoS) evaluation mechanism and the dynamic normal distribution selection method are adopted for service deployment. And the dynamic inertia particle swarm optimization (DI-PSO) algorithm is employed to realize task scheduling. Furthermore, the cloud platform and existing terminal server group are used to conduct platform structure comparison experiments, and the popular task scheduling algorithms are selected for simulation experiments. Experimental results of platform measurement show that the average service response time of different services can be improved by about 17.3% to 37.4%. The average occupancy ratio of computing resources can be reduced by about 5%. The simulation results show that the earliest completion time of single task list can be decreased by 11.3% to 20.9%, and the makespan of 100 task lists can be improved by 0.3 times.

In a public blockchain system applying Proof of Work (PoW), the participants need to compete with their computing resources for reward, which is challenging for resource-limited devices. Mobile blockchain is proposed to facilitate the application of blockchain for mobile service, in which the lightweight devices can participate mining by renting resources from the Cloud Computing Service Provider (CCSP), but CCSP usually does not have the information about the demand preference of users. In this paper, a contract model is adopted to address the cloud computing resource allocation and pricing problem in the mobile blockchain. In particular, an adverse selection contract solution is proposed to overcome the information asymmetry problem, and resource pooling is introduced to improve the stability of users rewards. Simulation results show that the information asymmetry problem is well overcome by adverse selection contract so that CCSP can obtain more utility than linear pricing contracts. Furthermore, the resource pooling could effectively improve the users and CCSPs utilities. When the size of the mining pool is large enough, it can achieve an improvement effect of more than 10 times. The effect of pool size and user type distribution on CCSPs utility is also studied.

Objective Breast cancer and breast tumors have been considered to be the most pervasive form of cancer in medical practice. Breast tumors are life-threatening to women, and their early detection could save lives with the proper treatment. Physical methods for detection of Breast Cancer are time-consuming and often prone to a misdiagnosis at classifying tumors. Recent trends in radiological imaging have significantly improved the efficiency and veracity of breast tumor classification. Artificial intelligence techniques could be used as an automated detection and classification system. Materials and methods In this research, we propose a novel configuration of a Stacking Ensemble with custom Convolutional Neural Network architectures to classify breast tumors from ultrasound images into ‘Normal’, ‘Benign’, and ‘Malignant’ categories. Results After thorough experimentation, our ensemble has performed with an accuracy, f1-score, precision, and recall of 92.15%, 92.21%, 92.26%, 92.17% respectively. Conclusion The presented ensemble leverages three Stacked Feature Extractors coupled with a characteristic meta-learner to provide an overall balanced classification performance, with better accuracy and lower false positives. The architecture works in association with gaussian dropout layers to improve the computation and an alternative pooling scheme to retain essential features.

Spatial downsampling layers are favored in convolutional neural networks (CNNs) to downscale feature maps for larger receptive fields and less memory consumption. However, for discriminative tasks, there is a possibility that these layers lose the discriminative details due to improper pooling strategies, which could hinder the learning process and eventually result in suboptimal models. In this paper, we present a unified framework over the existing downsampling layers (e.g., average pooling, max pooling, and strided convolution) from a local importance view. In this framework, we analyze the issues of these widely-used pooling layers and figure out the criteria for designing an effective downsampling layer. According to this analysis, we propose a conceptually simple, general, and effective pooling layer based on local importance modeling, termed as {\em Local Importance-based Pooling} (LIP). LIP can automatically enhance discriminative features during the downsampling procedure by learning adaptive importance weights based on inputs. Experiment results show that LIP consistently yields notable gains with different depths and different architectures on ImageNet classification. In the challenging MS COCO dataset, detectors with our LIP-ResNets as backbones obtain a consistent improvement ($\ge 1.4\%$) over the vanilla ResNets, and especially achieve the current state-of-the-art performance in detecting small objects under the single-scale testing scheme., Comment: ICCV 2019 camera ready version. Code: https://github.com/sebgao/LIP

Sequential recommendation aims to identify and recommend the next few items for a user that the user is most likely to purchase/review, given the user's purchase/rating trajectories. It becomes an effective tool to help users select favorite items from a variety of options. In this manuscript, we developed hybrid associations models (HAM) to generate sequential recommendations using three factors: 1) users long-term preferences, 2) sequential, high-order and low-order association patterns in the users most recent purchases/ratings, and 3) synergies among those items. HAM uses simplistic pooling to represent a set of items in the associations, and element-wise product to represent item synergies of arbitrary orders. We compared HAM models with the most recent, state-of-the-art methods on six public benchmark datasets in three different experimental settings. Our experimental results demonstrate that HAM models significantly outperform the state of the art in all the experimental settings, with an improvement as much as 46.6%. In addition, our run-time performance comparison in testing demonstrates that HAM models are much more efficient than the state-of-the-art methods, and are able to achieve significant speedup as much as 139.7 folds.

It is of great significance to apply deep learning for the early diagnosis of Alzheimer's disease (AD). In this work, a novel tensorizing GAN with high-order pooling is proposed to assess mild cognitive impairment (MCI) and AD. By tensorizing a three-player cooperative game-based framework, the proposed model can benefit from the structural information of the brain. By incorporating the high-order pooling scheme into the classifier, the proposed model can make full use of the second-order statistics of holistic magnetic resonance imaging (MRI). To the best of our knowledge, the proposed Tensor-train, High-order pooling and Semisupervised learning-based GAN (THS-GAN) is the first work to deal with classification on MR images for AD diagnosis. Extensive experimental results on Alzheimer's disease neuroimaging initiative (ADNI) data set are reported to demonstrate that the proposed THS-GAN achieves superior performance compared with existing methods, and to show that both tensor-train and high-order pooling can enhance classification performance. The visualization of generated samples also shows that the proposed model can generate plausible samples for semisupervised learning purpose.

Servitization is the activity of selling the services provided by the product rather than the product itself. It is a business model that might be environmentally superior to conventional selling. Servitization promises accessibility to the product’s functionality, pooling of consumer use, and potentially products of better design. However, it can also inflate consumption and result in a bigger environmental impact overall. In this paper, we compare servitization with traditional selling for a monopolist durable goods manufacturer from both an economic and environmental perspective. In this comparison, we define the durability of a product as the use capacity; that is, how many usages it can endure before reaching end of life. We study the firm’s durability decision, followed by the price/fee decision, and the consequent usage in the market under each model. We find that servitization produces durability levels that are robust to customer heterogeneity, and higher than selling. Overall, environmental superiority of servitization hinges on product related costs, customer heterogeneity, and market composition. It is, however, robust to varying environmental factors in the use and manufacturing phases. When we compare environmental preferability with the economic incentives of the firm, we observe that they are not always aligned.

Deep neural networks are the most used machine learning systems in the literature, for they are able to train huge amounts of data with a large number of parameters in a very effective way. However, one of the problems that such networks face is overfitting. There are many ways to address the overfitting issue, one of which is regularization using the dropout function. The use of dropout has the benefit of using a combination of different networks in one architecture and preventing units from co-adapting in an excessive way. The dropout function is known to work well in fully-connected layers as well as in pooling layers. In this work, we propose a novel method called Mixed-Pooling-Dropout that adapts the dropout function with a mixed-pooling strategy. The dropout operation is represented by a binary mask with each element drawn independently from a Bernoulli distribution. Experimental results show that our proposed method outperforms conventional pooling methods as well as the max-pooling-dropout method with an interesting margin (0.926 vs 0.868) regardless of the retaining probability.

Cloud computing provides many advantages services to its users like online resources accessibility, better cost management, dynamic resource pooling, and efficient Virtual Machine (VM) allocation. With so many useful services, cloud computing also introduces several security risks of cloud user information privacy as well as select the minimum number of VM to execute task load and improve resource utilization are some challenging issues. The objective of this work is to achieve higher resource utilization rate while considering the security of the cloud data. We extended the natural phenomena based Intelligent Water Drop (IWD) algorithm and proposed a VM allocation algorithm which optimizes the task execution in a secure cloud environment. The proposed work is implemented on the CloudSim simulation toolkit, and to test the validity of the algorithm, a comparison is performed with other well-known VM allocation policies in the cloud computing. The experimental simulation results showed that the proposed VM allocation policy performed better than exiting VM allocation approaches.

We propose the hybrid malware detection scheme, HyMalD, with bidirectional long short-term memory (Bi-LSTM) and the spatial pyramid pooling network (SPP-Net). Its purpose is to protect IoT devices and minimize the damage caused by infection through obfuscated malware. HyMalD performs the static and dynamic analyses logically simultaneously detects obfuscated malware, which is impossible to do using static analysis alone. First, it extracts static features of the opcode sequence using a reconstructed dataset according to the obfuscation, and extracts the application programming interface (API) call sequence dynamically. The extracted features are trained through the Bi-LSTM and SPP-Net models, which HyMalD uses to detect and classify IoT malware. The performance of HyMalD was evaluated, and its detection accuracy was 92.5%. False-negative rate (FNR) of HyMalD was 7.67%. Thus, HyMalD detects IoT malware more accurately and with a lower FNR than static analysis, which had a 92.09% detection accuracy and a 9.97% FNR.

In order to solve the problems of highly redundant spatial information and motion noise in the heart rate (HR) estimation from facial videos based on remote Photoplethysmography(rPPG), this paper proposes a novel HR estimation method based on spatial-temporal attention model. Firstly, to reduce the redundant information and strengthen the association relationships of long-range videos, the spatial-temporal facial features are extracted by the 2D convolutional neural network (2DCNN) and 3D convolutional neural network (3DCNN), respectively. The aggregation function is adopted to incorporate feature maps into short segment spatial-temporal feature maps. Secondly, the spatial-temporal strip pooling is designed in the spatial-temporal attention module to reduce head movement noises. Then, via the two-part loss function, the model can focus more on the rPPG signal rather than the interference. We conduct extensive experiments on two public datasets to verify the effectiveness of our model. The experimental results show that the proposed method achieves significantly better performances than the state-of-the-art baselines: The mean absolute error could be reduced by 11% on the PURE dataset, and by 25% on the COHFACE dataset.

Introduction: SARS-CoV-2 is a fatal disease of global public health concern. Measures to reduce its spread critically depend on timely and accurate diagnosis of virus-infected individuals. Biobanks can have a pivotal role in elucidating disease etiology, translation, and advancing public health. In this article, we show how a biobank has been a critical resource in the rapid response to coronavirus disease of 2019 (COVID-19) in Uganda. Materials and Methods: The Integrated Biorepository of H3Africa Uganda established a COVID-19 biobank. Standard Operating Procedures for sample and data collection, sample processing, and storage were developed. An e-questionnaire data tool was used to collect sociodemographic factors. Samples were collected at 7-day intervals from patients, analyzed for key parameters, processed, annotated, characterized, and stored at appropriate temperatures. Results: Stored samples have been used in validation of 17 diagnostic kits, the Cepheid Xpert Xpress SARS-CoV-2 assay, as well as a sample pooling technique for mass screening and polymerase chain reaction assay validation. Kits that passed validation were deployed for mass screening boosting early detection, isolation, and treatment of COVID-19 cases. Also, 10 applications from researchers and biotech companies have been received and approved and 4 grants have been awarded Conclusion: The CoV-Bank has proven to be an invaluable resource in the fight against the COVID-19 pandemic in Uganda, as samples have been resources in the validation and development of COVID-19 diagnostic tools, which are important in tracing and isolation of infected cases to confront, delay, and stop the spread of the SARS-CoV-2 virus.

Driver drowsiness is one of the main factors leading to road fatalities and hazards in the transportation industry. Electroencephalography (EEG) has been considered as one of the best physiological signals to detect drivers' drowsy states, since it directly measures neurophysiological activities in the brain. However, designing a calibration-free system for driver drowsiness detection with EEG is still a challenging task, as EEG suffers from serious mental and physical drifts across different subjects. In this paper, we propose a compact and interpretable Convolutional Neural Network (CNN) to discover shared EEG features across different subjects for driver drowsiness detection. We incorporate the Global Average Pooling (GAP) layer in the model structure, allowing the Class Activation Map (CAM) method to be used for localizing regions of the input signal that contribute most for classification. Results show that the proposed model can achieve an average accuracy of 73.22% on 11 subjects for 2-class cross-subject EEG signal classification, which is higher than conventional machine learning methods and other state-of-art deep learning methods. It is revealed by the visualization technique that the model has learned biologically explainable features, e.g., Alpha spindles and Theta burst, as evidence for the drowsy state. It is also interesting to see that the model uses artifacts that usually dominate the wakeful EEG, e.g., muscle artifacts and sensor drifts, to recognize the alert state. The proposed model illustrates a potential direction to use CNN models as a powerful tool to discover shared features related to different mental states across different subjects from EEG signals. National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore under its International Research Centres in Singapore Funding Initiative.

With the development of the imaging technology of various sensors, multisource image classification has become a key challenge in the field of image interpretation. In this article, a novel classification method, called the deep multiview union learning network (DMULN), is proposed to classify multisensor data. First, an associated feature extractor is designed to process the multisource data by canonical correlation analysis (CCA) in the head of the network. Second, an improved deep learning architecture with two branches is presented to extract high-level view features from the associated features. Third, a novel pooling, called view union pooling, is proposed to fuse the multiview feature from the deep model. Finally, the fused feature is fed into the classifier. The proposed framework is easy to optimize since it is an end-to-end network. Extensive experiments and analysis on the datasets IEEE_grss_dfc_2017 and IEEE_grss_dfc_2018 show that the proposed method achieves comparable results. Our results demonstrate that abundant multisource information can improve the classification performance.

This paper presents a novel heuristic method named Neural-Networks-Driven prediction (NEED) for robot path plan-ning problems. Different from classical heuristic methods, NEED is not designed for some specific environments, which can be applied to various environments. NEED has an encoder-decoder structure which consists of three modules: convolutional neural network (CNN) backbone, spatial pooling module and decoder module. By learning from a number of successful path planning cases, NEED can learn to analyse the environment structure and predict the promising search region for the new path planning problem. This predicted region serves as a heuristic to guide the search direction of path planning algorithms. A series of simulation experiments are conducted to demonstrate that NEED significantly improves the algorithm performance on solving new path planning problems. Meanwhile, NEED can be also applied to highly dynamic environments since it can output the prediction results at a speed of over 100 Hz on a laptop.

Clinician-reported outcome measures (ClinRO measures) play a fundamental role in quality assurance throughout healthcare systems. With commissioners turning ever more frequently to ClinRO data to evaluate and compare individual hospital performance and casemix, and funding decisions increasingly relying on these data, agreed core outcome sets (COS) are essential for the collection of standardised specialty-specific outcomes. Beyond their role in service commissioning, COS enable standardisation of outcomes in clinical studies, allowing comparisons to be drawn between similar trials as well as pooling of data for systematic reviews and metaanalyses. This review explores those ClinRO measures most commonly reported in the third molar literature, highlighting inconsistencies in ClinRO selection, measurement and reporting among researchers. We recognise here a prime opportunity for the specialty to address this lag in COS relative to other surgical specialties. With the Quality Outcomes in Maxillofacial Surgery (QOMS) overseeing the institution of many subspecialty-specific national databases in recent years, OMFS is well placed to develop a series of COS for each subspecialty domain for the benefit of researchers, clinicians and ultimately, patients.

In recent years, the rapid development of UAV technology has greatly improved the efficiency of the detection of concrete bridge cracks. With the increase in the number of bridge inspection UAVs, the number of tasks handled by cloud services has increased linearly, resulting in increased computational pressure on cloud services. In order to reduce the computational load of cloud servers, we proposed a crack segmentation network based on UAV-enabled edge computing. However, due to the limitation of computational capability of edge computing and the strength inhomogeneity and background complexity of cracks, crack detection is still a challenging task. Thus, we proposed an effective concrete crack segmentation network based on UAV-enabled edge computing, the network used feature map fusion to fuse different levels of feature map information into lower level features for crack detection. The atrous spatial pyramid pooling network was used to increase the low-resolution feature map receptive field information for cracks and to enhance the detection accuracy for cracks of different scales. In addition, loss functions for crack datasets were proposed to solve the problem of imbalance due to positive and negative samples in the concrete crack images. Experiments demonstrated that the proposed methods are better than the state-of-the-art edge detection and semantic segmentation methods in terms of accuracy and generality.

Adverse weather conditions seriously threaten the traffic safety, especially for rainy days with the ponding water on the road surface, which potentially result in vehicle crashes, person injuries and crash fatalities. Automatic splashed water detection based on surveillance videos is an attractive way to effectively prevent the traffic accidents. However, surveillance videos exhibit great variations with lighting changes, illumination conditions and complex backgrounds, which pose great difficulties in automatic recognition. In this paper, a novel deep learning based approach is proposed to detect the splashed water. To the best of our knowledge, this is the first work on this topic based on deep learning. An effective semantic segmentation network, called SWNet, is novelly proposed to extract the potential splashed water regions. An encoder-decoder structure is designed to capture the visual characteristics of splashed water. SWNet achieves high efficiency by reusing pooling indices and adopting the light-weight decoder. With the multi-scale feature fusion structure, SWNet integrates the coarse semantic information and detailed appearance information, which significantly boosts the accuracy and refines the edge segmentation. A weighted cross entropy loss for splashed water is adopted to cope with the unbalanced distribution between splashed water and backgrounds. Moreover, a splashed water attention module is designed to focus on the salient regions of moving vehicles and splashed water, by performing attention mechanism to integrate global contextual information in semantic segmentation. Experiments conducted on a newly collected splashed water dataset demonstrate the effectiveness and efficiency of the proposed approach, which outperforms the state-of-the-art methods.

Depression is a serious psychiatric disorder that restricts an individual's ability to work properly in both their daily and professional lives. Usually, the diagnosis of depression often needs a thorough assessment by an expert professional. Recently, significant consideration has been given to automatic depression prediction for more reliable and efficient depression investigation. In this paper, we propose a novel framework to estimate the depression level from video data by employing a two-stream deep spatiotemporal network. Our approach extracts spatial information using the Inception-ResNet-v2 network. In contrast, we introduce a volume local directional number (VLDN) based dynamic feature descriptor to capture facial motions. Then, the feature map obtained from the VLDN is fed into a convolutional neural network (CNN) to obtain more discriminative features. Additionally, we designed a multilayer bidirectional long short-term memory (Bi-LSTM) model to obtain temporal information by integrating the temporal median pooling (TMP) approach into the model. The TMP approach is employed on the temporal fragments of spatial and temporal features. Finally, extensive experimental analysis of two challenging datasets, AVEC2013 and AVEC2014, demonstrates that the proposed approach shows promising performance compared to the existing approaches for depression level prediction.

Due to the proliferation of sports-related multimedia content on the WWW, effective visual search and retrieval present interesting research challenges. These are caused by poor image quality, a wide range of possible camera points of view, pose variations on the part of athletes engaged in playing a sport, deformations of text appearing on sports person’s clothing and uniforms in motion, occlusions caused by other objects, etc. To address these challenges, this paper presents a new method for detecting text on human bodies in sports images. Unlike most existing methods, which attempt to exploit locations of a player’s torso, face, and skin, we propose an end-to-end episodic learning approach that employs inductive learning criteria for detecting clothing regions in an image, which are, in turn, then used for text detection. Our method integrates a Residual Network (ResNet) and Pyramidal Pooling Module (PPM) for generating a spatial attention map. The Progressive Scalable Expansion Algorithm (PSE) is adapted for text detection from these regions. Experimental results on our own dataset as well as several benchmarks (like RBNR and MMM which contain images of runners in marathons, and Re-ID which is a person re-identification dataset) demonstrate that the proposed method outperforms existing methods in terms of precision and F1-score. We also present results for sports images chosen from natural scene text detection datasets such as CTW1500 and MS-COCO to show the proposed method is effective and reliable across a range of inputs.

Congestion control has traditionally relied on monitoring packet-level performance (e.g. latency, loss) through feedback signals propagating end-to-end together with various queue management practices (e.g. carefully setting various parameters, such as router buffer thresholds) in order to regulate traffic flow. Due to its end-to-end nature, this approach is known to transfer data according to the path's slowest link, requiring several RTTs to transmit even a few tens of KB during slow start. In this paper, we take a radically different approach to control congestion, which obviates end-to-end performance monitoring and careful setting of network parameters. The resulting In-Network Resource Pooling Protocol (INRPP) extends the resource pooling principle to exploit in-network resources such as router storage and unused bandwidth along alternative sub-paths. In INRPP, content caches or large (possibly bloated) router buffers are used as a place of temporary custody for incoming data packets in a store and forward manner. Data senders push data in the network and when it hits the bottleneck link, in-network caches at every hop store data in excess of the link capacity; nodes progressively move/send data (from one cache to the next) towards the destination. At the same time alternative sub-paths are exploited to move data faster towards the destination. We demonstrate through extensive simulations that INRPP is TCP friendly, and improves flow completion time and fairness by as much as 50% compared to RCP, MPTCP and TCP, under realistic network conditions.

Recognizing and organizing different series in an MRI examination is important both for clinical review and research, but it is poorly addressed by the current generation of picture archiving and communication systems (PACSs) and post-processing workstations. In this paper, we study the problem of using deep convolutional neural networks for automatic classification of abdominal MRI series to one of many series types. Our contributions are three-fold. First, we created a large abdominal MRI dataset containing 3717 MRI series including 188,665 individual images, derived from liver examinations. 30 different series types are represented in this dataset. The dataset was annotated by consensus readings from two radiologists. Both the MRIs and the annotations were made publicly available. Second, we proposed a 3D pyramid pooling network, which can elegantly handle abdominal MRI series with varied sizes of each dimension, and achieved state-of-the-art classification performance. Third, we performed the first ever comparison between the algorithm and the radiologists on an additional dataset and had several meaningful findings.

Despite the significant advances in iris segmentation, accomplishing accurate iris segmentation and localization in non-cooperative environment with visible illumination remains a grand challenge. In this paper, we present an end-to-end multi-task deep neural network, referred to as Iris R-CNN, to achieve superior accuracy for iris segmentation and localization. Inspired by instance segmentation, Iris R-CNN seamlessly integrates segmentation and localization in a unified framework and generates a normalized iris image/mask required for iris recognition. It is composed of several novel techniques to carefully explore the unique characteristics of iris. First, we propose two novel networks, (i) Double-Circle Region Proposal Network (DC-RPN) and (ii) Double-Circle Classification and Regression Network (DC-CRN), to efficiently capture pupil and iris circles and enhance the accuracy for iris localization. Second, we propose a novel normalization scheme for Regions of Interest (RoIs) to enable a radically new pooling operation over a double-circle region. To facilitate accurate training and validation, we annotate two public datasets in non-cooperative environment with visible illumination: NICE-II and MICHE. Experimental results on these two challenging datasets demonstrate the superior accuracy of our proposed approach over other state-of-the-art methods.

Both traditional retailers and e-tailers have been implementing omnichannel strategies such as buy online, pick up at store (BOPS). We build a stylized model to investigate the impact of the BOPS initiative on store operations from an inventory perspective. We consider two segments of customers, namely store-only customers who only make purchases offline and omni-customers who strategically choose between offline and online channels. We show that BOPS may either benefit or hurt the retailer depending on two fundamental system primitives: the store visiting cost and the online waiting cost. If the online waiting cost is relatively low and the store visiting cost is even lower, BOPS can induce omni-customers to migrate from online buying to BOPS, leading to demand pooling at the brick-and-mortar (B&M) store. Such demand pooling provides two benefits for the retailer: it reduces the overstocking cost, and after inventory reoptimization, it results in a higher fill rate at the B&M store, which benefits existing customers and potentially attracts more customers to the store. In contrast, if both store visiting and online waiting costs are relatively high with the latter even higher, introducing BOPS can result in demand depooling as a result of the migration of the omni-customers from offline purchasing to BOPS. This leads to a lower fill rate after inventory reoptimization, likely the result of a lower profit margin under BOPS, which turns away store-only customers and hurts the retailer. This paper was accepted by Charles Corbett, operations management.

The sustainable development of marine fisheries depends on the accurate measurement of data on fish stocks. Semantic segmentation methods based on deep learning can be applied to automatically obtain segmentation masks of fish in images to obtain measurement data. However, general semantic segmentation methods cannot accurately segment fish objects in underwater images. In this study, a Dual Pooling‐aggregated Attention Network (DPANet) to adaptively capture long‐range dependencies through an efficient and computing‐friendly manner to enhance feature representation and improve segmentation performance is proposed. Specifically, a novel pooling‐aggregate position attention module and a pooling‐aggregate channel attention module are designed to aggregate contexts in the spatial dimension and channel dimension, respectively. These two modules adopt pooling operations along the channel dimension and along the spatial dimension to aggregate information, respectively, thus reducing computational costs. In these modules, attention maps are generated by four different paths and are aggregated into one. The authors conduct extensive experiments to validate the effectiveness of the DPANet and achieve new state‐of‐the‐art segmentation performance on the well‐known fish image dataset DeepFish as well as on the underwater image dataset SUIM, achieving a Mean IoU score of 91.08% and 85.39% respectively, while significantly reducing FLOPs of attention modules by about 93%.

The rapid spread of the novel coronavirus disease 2019 (COVID‐19) causes a significant impact on public health. It is critical to diagnose COVID‐19 patients so that they can receive reasonable treatments quickly. The doctors can obtain a precise estimate of the infection's progression and decide more effective treatment options by segmenting the CT images of COVID‐19 patients. However, it is challenging to segment infected regions in CT slices because the infected regions are multi‐scale, and the boundary is not clear due to the low contrast between the infected area and the normal area. In this paper, a coarse‐refine segmentation network is proposed to address these challenges. The coarse‐refine architecture and hybrid loss is used to guide the model to predict the delicate structures with clear boundaries to address the problem of unclear boundaries. The atrous spatial pyramid pooling module in the network is added to improve the performance in detecting infected regions with different scales. Experimental results show that the model in the segmentation of COVID‐19 CT images outperforms other familiar medical segmentation models, enabling the doctor to get a more accurate estimate on the progression of the infection and thus can provide more reasonable treatment options.

Recently, the social media podium used to share and post patient medical content, which significantly lead to mining medical content for pharmacovigilance, drug repositioning, and Healthcare sector. Paramedical advertisers extract trusted health information from the social media platform and correlate health care information with the user profile for recommending proper treatment. The mining of the patient's personal and medical information on social media makes the patient aware of the recent drug revolution. For predicting computational medical services, there is a need to Develop a centralized Social media-based pharmacovigilance activities virtual podium that identify drug and ADR associated with it. Build an informational community of homogenous E-patient over the virtual platform for building informational and emotional support more efficiently. In this paper, a comprehensive overview of the recommendation system and possible research gap associated with drug repository and ADR are presented based on recent research. Furthermore, this paper present clinical post pre-processing, clinical entity recognition, tweet pooling and expert rating framework to classify Disease-Drug-ADR intersection for recommending drug and ADR. Simultaneously this framework present comparative analysis of benchmark classifier to trained and label corpus of tweets for drug repositioning and ADR automatically.

Recent graph neural network (GNN) based methods for few-shot learning (FSL) represent the samples of interest as a fully-connected graph and conduct reasoning on the nodes flatly, which ignores the hierarchical correlations among nodes. However, real-world categories may have hierarchical structures, and for FSL, it is important to extract the distinguishing features of the categories from individual samples. To explore this, we propose a novel hierarchical graph neural network (HGNN) for FSL, which consists of three parts, i.e., bottom-up reasoning, top-down reasoning, and skip connections, to enable the efficient learning of multi-level relationships. For the bottom-up reasoning, we design intra-class k-nearest neighbor pooling (intra-class knnPool) and inter-class knnPool layers, to conduct hierarchical learning for both the intra- and inter-class nodes. For the top-down reasoning, we propose to utilize graph unpooling (gUnpool) layers to restore the down-sampled graph into its original size. Skip connections are proposed to fuse multi-level features for the final node classification. The parameters of HGNN are learned by episodic training with the signal of node losses, which aims to train a well-generalizable model for recognizing unseen classes with few labeled data. Experimental results on benchmark datasets have demonstrated that HGNN outperforms other state-of-the-art GNN based methods significantly, for both transductive and non-transductive FSL tasks. The dataset as well as the source code can be downloaded online.

This research article proposes an automatic frame work for detecting COVID -19 at the early stage using chest X-ray image. It is an undeniable fact that coronovirus is a serious disease but the early detection of the virus present in human bodies can save lives. In recent times, there are somany research solutions that have been presented for early detection, but there is still a lack in need of right and even rich technology for its early detection. The proposed deep learning model analysis the pixels of every image and adjudges the presence of virus. The classifier is designed in such a way so that, it automatically detects the virus present in lungs using chest image. This approach uses an image texture analysis technique called granulometric mathematical model. Selected features are heuristically processed for optimization using novel multi scaling deep learning called light weight residual-atrous spatial pyramid pooling (LightRES-ASPP-Unet) Unet model. The proposed deep LightRES-ASPPUnet technique has a higher level of contracting solution by extracting major level of image features. Moreover, the corona virus has been detected using high resolution output. In the framework, atrous spatial pyramid pooling (ASPP) method is employed at its bottom level for incorporating the deep multi scale features in to the discriminative mode. The architectural working starts from the selecting the features from the image using granulometric mathematical model and the selected features are optimized using LightRESASPP- Unet. ASPP in the analysis of images has performed better than the existing Unet model. The proposed algorithm has achieved 99.6% of accuracy in detecting the virus at its early stage. © 2022 Tech Science Press. All rights reserved.

Hierarchical deep features can provide multilevel abstractions of target objects, which play an important role in target localization and classification. How to effectively aggregate abstract information from different levels in RGB and thermal modal is the key to exploit their complementary advantages for robust RGBT tracking. However, existing RGBT tracking algorithms either focus on the semantic information of the last layer or aggregate the hierarchical deep features from each modal by simple operations (e.g., summation and concatenation), which limit the capability of multimodal tracker. To address these issues, in this paper, we propose a novel multimodal cross-layer bilinear pooling network for RGBT tracking. In our network, firstly, to further boost the performance of the tracker, we adopt the channel attention mechanism to implement the adaptive calibration of feature channels for all convolutional layer features before realizing the hierarchical feature fusion. Then, we perform the bilinear pooling operation on any two layers through the cross product, which is a second-order computation and aggregates the deep semantic information and shallow texture information of target effectively. Finally, we design a quality-aware fusion module to aggregate bilinear pooling features of different layer interactions between different modalities in an adaptive manner. A large number of experiments on two public benchmark datasets demonstrate the effectiveness of our tracker against other state-of-the-art tracking methods.

Omnidirectional video, also known as 360-degree video, has become increasingly popular nowadays due to its ability to provide immersive and interactive visual experiences. However, the ultra high resolution and the spherical observation space brought by the large spherical viewing range make omnidi-rectional video distinctly different from traditional 2D video. To date, the video quality assessment (VQA) for omnidirectional video is still an open issue. The existing VQA metrics for omnidirectional video only consider the spatial characteristics of distortions, but the temporal changes of spatial distortions can also considerably influence human visual perception. In this paper, we propose a spatiotemporal modeling approach to evaluate the quality of the omnidirectional video. Firstly, we construct a spatiotemporal quality assessment unit to evaluate the average distortion in temporal dimension at the eye fixation level, based upon which the smoothed distortion value is recursively calculated and consolidated by the characteristics of temporal variations. Then, we give a detailed solution of how to to integrate the three existing spatial VQA metrics into our approach. Besides, the cross-format omnidirectional video distortion measurement is also investigated. Finally, the spatiotemporal distortion of the whole video sequence is obtained by pooling. Based on the modeling approach, a full reference objective quality assessment metric for omnidirectional video is derived, namely OV-PSNR. The experimental results show that our proposed OV-PSNR greatly improves the prediction performance of the existing VQA metrics for omnidirectional video.

Salient object detection (SOD) in optical remote sense images (ORSIs) is a valuable and challenging task. The factors in ORSI, such as background clutter, lighting shadows, imaging blur, and low resolution, significantly degrade the completeness and accuracy of salient objects. To handle this problem, we propose a novel model to learn robust multiscale region features of salient objects by simultaneously optimizing their boundaries. First, we extract multiscale region features of salient objects through a hierarchical attention module. Second, we generate the boundary features by combining the local cues and the global information generated by pyramid pooling. Finally, we embed the boundary features into region features at multiple scales. In particular, we design a joint learning scheme based on a bidirectional feature transformation to optimize boundary and region features simultaneously for accurate ORSI SOD. To provide a comprehensive evaluation platform, we construct a new dataset called ORSI-4199 for ORSI SOD. It contains 4199 finely annotated image pairs with diverse scenes, in which nine attributes (i.e., challenge types) are annotated to facilitate analyzing the strengths and weaknesses of SOD models from different perspectives. Extensive experiments on the public dataset ORSSD, EORRSD, and the newly created dataset ORSI-4199 show that the proposed approach achieves promising results against state-of-the-art methods. https://github.com/wchao1213/ORSI-SOD.

Public speaking is a critical skill in daily communication. While more practicing such as rehearsal is helpful to improve such a skill, lack of personalized feedback limits the effectiveness of practicing. Therefore, we formulate the task of personalized feedback as an affective audio annotation problem by learning knowledge from online public speech videos. Considering the great success of deep learning techniques such as convolutional neural networks in a wide range of applications including speech recognition and object recognition, we propose a novel convolutional clustering neural network (CCNN) to solve this multi-label classification problem. Instead of aggregating the features of different channels through pooling, we introduce a novel clustering layer to derive intermediate representation for improved annotation performance. In order to evaluate the performance of our proposed method, we purposely built an affective audio annotation dataset by collecting more than 2,000 video clips from the TED website. Experimental results on this dataset demonstrate that our proposed method outperforms traditional CNN-based approaches with a lower hamming loss for affective annotation.

Most pooling methods in state-of-the-art speaker embedding networks are implemented in the temporal domain. However, due to the high non-stationarity in the feature maps produced from the last frame-level layer, it is not advantageous to use the global statistics (e.g., means and standard deviations) of the temporal feature maps as aggregated embeddings. This motivates us to explore stationary spectral representations and perform aggregation in the spectral domain. In this paper, we propose attentive short-time spectral pooling (attentive STSP) from a Fourier perspective to exploit the local stationarity of the feature maps. In attentive STSP, for each utterance, we compute the spectral representations through a weighted average of the windowed segments within each spectrogram by attention weights and aggregate their lowest spectral components to form the speaker embedding. Because most energy of the feature maps is concentrated in the low-frequency region in the spectral domain, attentive STSP facilitates the information aggregation by retaining the low spectral components only. Moreover, due to the segment-level attention mechanism, attentive STSP can produce smoother attention weights (weights with less variations) than attentive pooling and generalize better to unseen data, making it more robust against the adverse effect of the non-stationarity in the feature maps. Attentive STSP is shown to consistently outperform attentive pooling on VoxCeleb1, VOiCES19-eval, SRE16-eval, and SRE18-CMN2-eval. This observation suggests that applying segment-level attention and leveraging low spectral components can produce discriminative speaker embeddings.

Limited by the shape-fixed kernels, convolutional neural networks (CNNs) are usually difficult to model difform land covers in hyperspectral images (HSIs), leading to inadequate land use. Recently, benefiting from the ability to conduct shape-adaptive convolutions and model complex patterns in graph-structured data, graph convolutional networks (GCNs) have been applied to HSI classification. However, due to the massive computation in GCNs, HSI is usually pretreated into a graph based on a specific superpixel segmentation, which limits the modeling of spatial topologies to the same scale. To break this limitation, we propose a multilevel superpixel structured graph U-Net (MSSGU) to learn multiscale features on multilevel graphs. Specifically, we construct several hierarchical segmentations from fine to coarse by progressively merging adjacent superpixels and then convert them into multilevel graphs. Meanwhile, based on the merging relations between hierarchical superpixels, we establish the pooling and unpooling functions to transfer features from one graph to another, thereby enabling different-level graphs to collaborate in a single network. Different from concatenating different-scale features straightforwardly in the feature fusion stage, MSSGU fuses them in a coarse-to-fine progressive manner, which can generate subtler fusion features adaptive to the pixelwise classification task. Moreover, we use a CNN instead of GCN to extract and fuse the pixel-level features, which greatly reduces the computation. Such a hybrid U-Net can exploit features of HSIs from a multiscale hierarchical perspective, and its performance has been proven competitive with other deep-learning-based methods by extensive experiments on three benchmark datasets.

Seismic fault detection is indispensable for exploring reservoirs of hydrocarbons, and a considerable amount of research has thus been devoted to it. With the rapid development of deep learning in recent years, researchers have begun using convolutional neural networks (CNNs) to identify seismic faults. However, it remains challenging for geophysical interpreters to quickly train an efficient model on a limited number of samples for fault interpretation. In this letter, we propose a workflow that uses a CNN-based method of semantic segmentation to interpret faults by using a small training set. It requires only the selection of some 2-D seismic sections from seismic volume data for interpretation and labeling to be trained to predict faults in the entire area considered. We simplified the VGG16 model to reduce training time and modified it to improve its performance. We used convolution layers instead of fully connected layers at the end of the network to implement the end-to-end classification of seismic images, and adopted dilation convolution to increase the receptive field and hybrid dilation convolution to avoid problems in it. We also applied the atrous spatial pyramid pooling (ASPP) module to further enhance the effects of segmentation. The results were subjected to postprocessing for refinement. The promising performance of the proposed method was verified on a set of real seismic data.

We introduce a simple yet effective network that embeds a novel Discriminative Feature Pooling (DFP) mechanism and a novel Video Segment Attention Model (VSAM), for video-based human action recognition from both trimmed and untrimmed videos. Our DFP module introduces an attentional pooling mechanism for 3D Convolutional Neural Networks that attentionally pools 3D convolutional feature maps to emphasize the most critical spatial, temporal, and channel-wise features related to the actions within a video segment, while our VSAM ensembles these most critical features from all video segments and learns (1) class-specific attention weights to classify the video segments into the corresponding action categories, and (2) class-agnostic attention weights to rank the video segments based on their relevance to the action class. Our action recognition network can be trained from both trimmed videos in a fully-supervised way and untrimmed videos in a weakly-supervised way. For untrimmed videos with weak labels, our network learns attention weights without the requirement of precise temporal annotations of action occurrences in videos. Evaluated on the untrimmed video datasets of THUMOS14 and ActivityNet1.2, and trimmed video datasets of HMDB51, UCF101, and HOLLYWOOD2, our network achieves superior performance, compared to the latest state-of-the-art methods

Land cover classification is the basic task of remote sensing image interpretation. Related methods have developed rapidly, especially the branch based on deep learning (DL). For high-resolution remote sensing images, the smaller inter-class difference and greater intra-class difference are two obstacles to improving the classification accuracy. For the former, the DL models generally use a deeper encoder to extract more powerful classification features. Considering that the scale of different land cover categories varies greatly, multi-scale feature extraction modules are also used to improve the classification accuracy. While the latter is always overlooked, and thus we propose a dual-output model, which uses a dense spatial pyramid pooling (DSPP) module to generate both the pixel-level and region-level predictions, to reduce the influence of intra-class differences. To further increase the classification accuracy, we investigate the band selection technique to apply the pre-trained encoder from the natural red green blue (RGB) dataset to multi-spectral remote sensing images. Extensive experiments on two datasets demonstrate the effectiveness of our model.

Point set registration is one of the challenging tasks in remote sensing image processing and analysis. Its critical step is to find the corresponding relationships between the fixed scene point set and the moving model point set that undergo different sorts of transformations. Existing algorithms primarily utilize different types of prior information to improve the registration performance, such as spatial consistency, local similarity, and uniform outliers. However, due to the lack of active evaluation on the prior and intermediate information during the registration process, these strategies are susceptible to large data transformations. In order to enhance the robustness and accuracy for point set registration, we propose in this article a novel framework, namely Registration-is-Evaluation (RisE). Based on a multigranular probability model, our method exploits and utilizes both prior and posterior information to dynamically evaluate the matching status. What is more, instead of adding an extra uniform prior, we unified the outliers, noise, missing points, and heavily warped points into our registration evaluation model and address them simultaneously. We also apply a novel point set descriptor, called local polar relative geometry (LPRG), to have a more robust local similarity measurement. It adopts the local polar coordinate to perform multiscale pooling and relative geometric computation. Based on our proposed method, the matching relationships and the spatial transformations can be actively evaluated to provide useful contextual guidance for the registration process. Experimental results on multiple data sets show that our algorithm outperforms the state-of-the-art methods, in terms of both accuracy and robustness under large data degradations.

Multi-view object detection is challenging due to the influence of the different view-angles on intra-class similarity. The uniformed feature representation of traditional detectors couples the object’s perspective attribute and semantic feature, and the variances of perspective will cause intra-class differences. In this paper, a robust perspective-sensitive network (PSNet) is proposed to overcome the above problem. The uniformed feature is replaced by the perspective-specific structural feature, which makes the network perspective sensitive. Its essence is to learn multiple perspective spaces. In each perspective space, the semantic feature is decoupled from the perspective attribute and is robust to perspective variances. Perspective-sensitive RoI pooling and loss function are proposed for perspective-sensitive learning. Experiments on Pascal3D + and SpaceNet MOVI show the effectiveness and superiority of the PSNet.

AIM: Alcoholism is a disease that a patient becomes dependent or addicted to alcohol. This paper aims to design a novel artificial intelligence model that can recognize alcoholism more accurately. METHODS: We propose the VGG-Inspired stochastic pooling neural network (VISPNN) model based on three components: (i) a VGG-inspired mainstay network, (ii) the stochastic pooling technique, which aims to outperform traditional max pooling and average pooling, and (iii) an improved 20-way data augmentation (Gaussian noise, salt-and-pepper noise, speckle noise, Poisson noise, horizontal shear, vertical shear, rotation, Gamma correction, random translation, and scaling on both raw image and its horizontally mirrored image). In addition, two networks (Net-I and Net-II) are proposed in ablation studies. Net-I is based on VISPNN by replacing stochastic pooling with ordinary max pooling. Net-II removes the 20-way data augmentation. RESULTS: The results by ten runs of 10-fold cross-validation show that our VISPNN model gains a sensitivity of 97.98±1.32, a specificity of 97.80±1.35, a precision of 97.78±1.35, an accuracy of 97.89±1.11, an F1 score of 97.87±1.12, an MCC of 95.79±2.22, an FMI of 97.88±1.12, and an AUC of 0.9849, respectively. CONCLUSION: The performance of our VISPNN model is better than two internal networks (Net-I and Net-II) and ten state-of-the-art alcoholism recognition methods.

Deep learning algorithms have been widely utilized for synthetic aperture radar (SAR) target detection. Nevertheless, the traditional feature extraction methods and deep learning methods achieve improved ship detection accuracy at a cost of increased complexity and lower detection speed. As detection speed also is meaningful, especially in real-time maritime rescue and emergency military decision-making applications, we propose a new framework of faster region-based convolutional neural network (R-CNN) detection method to handle this problem. A new lightweight basic network with feature relay amplification and multiscale feature jump connection structure is designed to extract the features of each scale target in the SAR images, so as to improve its recognition and localization task network. Moreover, the K-Means method is used to obtain the distribution of the target scale, which enables to select more appropriate preset anchor boxes to reduce the difficulty of network learning. Finally, RoIAlign instead of region of interest (RoI) Pooling is used to reduce the quantization error during positioning. Experimental results show that the detection performance of the proposed method achieves 0.898 average precision (AP), which is 2.78% better than the conventional Faster R-CNN and 800% faster detection speed.

This paper analyzes consumers' privacy choice concerning their data and firms' ensuing pricing strategies. Consumers decide whether to reveal private information in form of cookies. We study this endogenous decision in a duopoly model with behavior-based pricing and conduct an experiment to test our theoretical predictions. Contrasting two data environments, we derive a unique pure-strategy pooling equilibrium for each environment. We find that all consumers share their data in an open data environment while no consumers share data in an exclusive environment. In the experiment, however, some subjects readily share their data in the exclusive treatment. In the open data treatment subjects predominantly act as predicted by our theory. A mandated data sharing policy among firms leads consumers to share more data and increases consumer welfare. This helps policy makers to make an informed decision about data protection policies.

Population testing for severe acute respiratory syndrome coronavirus 2 is necessary because of the potential for viral transmission from asymptomatic cases, yet the scarcity of reagents and equipment has increased the cost-prohibitive implementation of screening campaigns at institutions of higher education. Significant analytical sensitivities of nucleic acid amplification methods permit sample pooling to increase testing capacity. Statistical models compared optimal testing configurations for pools of 3, 5, and 10 samples. Assessment of pooling using the TaqPath COVID-19 Combo Kit multiplex assay (ORF1ab, N, and S gene targets) involved a limit-of-detection study, matrix-effect study, and clinical comparison of neat with pooled samples. A limit of detection of 135.02 (ORF1ab; 95% CI, 117.21–155.52), 373.92 (N; 95% CI, 257.05–437.64), and 1001.32 (S; 95% CI, 896.62–1118.33) gene copy equivalents per milliliter was resolved. Seventy-two randomly selected samples showed slight suppression owing to a negative sample matrix. The resulting mean cycle threshold shifts were 2.09 (ORF1ab), 1.76 (N), and 2.31 (S) for the 3-sample pool, 2.83 (ORF1ab), 2.45 (N), and 3.24 (S) for the 5-sample pool, and 3.99 (ORF1ab), 3.46 (N), and 4.07 (S) for the 10-sample pool. Despite a quantitative sensitivity loss trend, the qualitative result was unaffected in each pool. According to the range of disease prevalence observed at the testing site (0.03% to 7.32%), a pool of five samples was deemed an optimal and cost-effective option for monitoring the Northeastern University community.

Multi-modal fusion has been proved to help enhance the performance of scene classification tasks. This paper presents a 2D-3D Fusion stage that combines 3D Geometric Features with 2D Texture Features obtained by 2D Convolutional Neural Networks. To get a robust 3D Geometric embedding, a network that uses two novel layers is proposed. The first layer, Multi-Neighbourhood Graph Convolution, aims to learn a more robust geometric descriptor of the scene combining two different neighbourhoods: one in the Euclidean space and the other in the Feature space. The second proposed layer, Nearest Voxel Pooling, improves the performance of the well-known Voxel Pooling. Experimental results, using NYU-Depth-V2 and SUN RGB-D datasets, show that the proposed method outperforms the current state-of-the-art in RGB-D indoor scene classification task., Comment: Information Fusion 2021

The popularity of smartphones and the advent of GPS positioning and wireless communication technologies in the recent decade have facilitated large-scale implementations of dynamic ridepooling services, such as Uber Pool, Lyft Line, and Didi Pinche. As in such services trips usually start before the appearance of pooling partners, knowing the probability of getting matching with another order (i.e., matching probability), the expected detour distance, and the expected shared distance before the start of each trip is essential for passengers to evaluate their willingness to pool and for ridepooling platforms to offer attractive discounts. In this paper, assuming that every ridepooling passenger shares vehicle space with at most one another during the entire trip, and ridepooling orders in each (origin-destination) OD pair appear following a Poisson process with a given rate, we propose a mathematical modeling approach to predict the matching probability, the expected ride distance, and the expected shared distance of each order under a first-come-first-serve strategy in dynamic ridepooling service. The method defines unmatched passengers at different locations along the exclusive-riding path of each OD pair into different seeker- and taker-states, formulates the complex interdependency of the matching probabilities, matching rates and arrival rates of (unmatched) passengers in different states into a system of nonlinear equations, and generates the matching probabilities and expected ride/shared distances of all OD pairs simultaneously. Under the same first-come-first-serve strategy, we simulated the occurrence, movements and state transitions of ridepooling orders based on a 30*30 grid network and the real network of Haikou City in China. In comparison with simulation results, we show that the method proposed in this paper can generate fairly satisfactory predictions under diverse matching conditions and demand intensities.

The analysis of the structure of musical pieces is a task that remains a challenge for Artificial Intelligence, especially in the field of Deep Learning. It requires prior identification of the structural boundaries of the music pieces, whose structural boundary analysis has recently been studied with unsupervised methods and supervised neural networks trained with human annotations. The supervised neural networks that have been used in previous studies are Convolutional Neural Networks (CNN) that use Mel-Scaled Log-magnitude Spectograms features (MLS), Self-Similarity Matrices (SSM) or Self-Similarity Lag Matrices (SSLM) as inputs. In previously published studies, pre-processing is done in different ways using different distance metrics, and different audio features are used for computing the inputs, so a generalised pre-processing method for calculating model inputs is missing. The objective of this work is to establish a general method to pre-process these inputs by comparing the results obtained by taking the inputs calculated from different pooling strategies, distance metrics and audio characteristics, also taking into account the computing time to obtain them. We also establish the most effective combination of inputs to be delivered to the CNN to provide the most efficient way to extract the boundaries of the structure of the music pieces. With an adequate combination of input matrices and pooling strategies, we obtain an accuracy F1 of 0.411 that outperforms a current work done under the same conditions (same public available dataset for training and testing).

Using single-scale prediction kernels or Region of Interest (RoI) pooling in the prediction modules of modern object detectors is not very successful in matching different scales of objects. State-of-the-art detectors with the feature pyramid structure built on different resolutions of feature maps can help alleviate this problem. Although with this structure, single-scale prediction kernels or RoI pooling still struggles to detect small objects, and simultaneously, the former continues to encounter the misalignment problem on very large objects. In this paper, we propose the attention-guided pyramidal prediction kernels module with a customized IoU-adaptive loss function to deal with the misalignment problem between the prediction module and different scales of objects. To mitigate the effect of heavy detection head, we also introduce the salient object regions recognition module to identify these regions that have strong object cues. Additionally, interleaved subsampling, as the proposed feature enhancement approach, is used to generate highly discriminative feature representations. We refer to the detection framework constituted by these proposed methods as SnipeDet. Results show that SnipeDet achieves 41.1 AP at the speed of 15.4 FPS on the MS COCO test-dev set with 512 × 512 input images, which outperforms state-of-the-art one-stage detectors and has a better trade-off between speed and accuracy.