In recent years,audio pattern recognition has emerged as a key area of research,driven by its applications in human-computer interaction,robotics,and healthcare.Traditional methods,which rely heavily on handcrafted fe...In recent years,audio pattern recognition has emerged as a key area of research,driven by its applications in human-computer interaction,robotics,and healthcare.Traditional methods,which rely heavily on handcrafted features such asMel filters,often suffer frominformation loss and limited feature representation capabilities.To address these limitations,this study proposes an innovative end-to-end audio pattern recognition framework that directly processes raw audio signals,preserving original information and extracting effective classification features.The proposed framework utilizes a dual-branch architecture:a global refinement module that retains channel and temporal details and a multi-scale embedding module that captures high-level semantic information.Additionally,a guided fusion module integrates complementary features from both branches,ensuring a comprehensive representation of audio data.Specifically,the multi-scale audio context embedding module is designed to effectively extract spatiotemporal dependencies,while the global refinement module aggregates multi-scale channel and temporal cues for enhanced modeling.The guided fusion module leverages these features to achieve efficient integration of complementary information,resulting in improved classification accuracy.Experimental results demonstrate the model’s superior performance on multiple datasets,including ESC-50,UrbanSound8K,RAVDESS,and CREMA-D,with classification accuracies of 93.25%,90.91%,92.36%,and 70.50%,respectively.These results highlight the robustness and effectiveness of the proposed framework,which significantly outperforms existing approaches.By addressing critical challenges such as information loss and limited feature representation,thiswork provides newinsights and methodologies for advancing audio classification and multimodal interaction systems.展开更多
Machine-to-machine (M2M) communication networks consist of resource-constrained autonomous devices, also known as autonomous Internet of things (IoTs) or machine-type communication devices (MTCDs) which act as a backb...Machine-to-machine (M2M) communication networks consist of resource-constrained autonomous devices, also known as autonomous Internet of things (IoTs) or machine-type communication devices (MTCDs) which act as a backbone for Industrial IoT, smart cities, and other autonomous systems. Due to the limited computing and memory capacity, these devices cannot maintain strong security if conventional security methods are applied such as heavy encryption. This article proposed a novel lightweight mutual authentication scheme including elliptic curve cryptography (ECC) driven end-to-end encryption through curve25519 such as (i): efficient end-to-end encrypted communication with pre-calculation strategy using curve25519;and (ii): elliptic curve Diffie-Hellman (ECDH) based mutual authentication technique through a novel lightweight hash function. The proposed scheme attempts to efficiently counter all known perception layer security threats. Moreover, the pre-calculated key generation strategy resulted in cost-effective encryption with 192-bit curve security. It showed comparative efficiency in key strength, and curve strength compared with similar authentication schemes in terms of computational and memory cost, communication performance and encryption robustness.展开更多
To capture the nonlinear dynamics and gain evolution in chirped pulse amplification(CPA)systems,the split-step Fourier method and the fourth-order Runge–Kutta method are integrated to iteratively address the generali...To capture the nonlinear dynamics and gain evolution in chirped pulse amplification(CPA)systems,the split-step Fourier method and the fourth-order Runge–Kutta method are integrated to iteratively address the generalized nonlinear Schrödinger equation and the rate equations.However,this approach is burdened by substantial computational demands,resulting in significant time expenditures.In the context of intelligent laser optimization and inverse design,the necessity for numerous simulations further exacerbates this issue,highlighting the need for fast and accurate simulation methodologies.Here,we introduce an end-to-end model augmented with active learning(E2E-AL)with decent generalization through different dedicated embedding methods over various parameters.On an identical computational platform,the artificial intelligence–driven model is 2000 times faster than the conventional simulation method.Benefiting from the active learning strategy,the E2E-AL model achieves decent precision with only two-thirds of the training samples compared with the case without such a strategy.Furthermore,we demonstrate a multi-objective inverse design of the CPA systems enabled by the E2E-AL model.The E2E-AL framework manifests the potential of becoming a standard approach for the rapid and accurate modeling of ultrafast lasers and is readily extended to simulate other complex systems.展开更多
The primary challenge in weakly supervised semantic segmentation is effectively leveraging weak annotations while minimizing the performance gap compared to fully supervised methods.End-to-end model designs have gaine...The primary challenge in weakly supervised semantic segmentation is effectively leveraging weak annotations while minimizing the performance gap compared to fully supervised methods.End-to-end model designs have gained significant attention for improving training efficiency.Most current algorithms rely on Convolutional Neural Networks(CNNs)for feature extraction.Although CNNs are proficient at capturing local features,they often struggle with global context,leading to incomplete and false Class Activation Mapping(CAM).To address these limitations,this work proposes a Contextual Prototype-Based End-to-End Weakly Supervised Semantic Segmentation(CPEWS)model,which improves feature extraction by utilizing the Vision Transformer(ViT).By incorporating its intermediate feature layers to preserve semantic information,this work introduces the Intermediate Supervised Module(ISM)to supervise the final layer’s output,reducing boundary ambiguity and mitigating issues related to incomplete activation.Additionally,the Contextual Prototype Module(CPM)generates class-specific prototypes,while the proposed Prototype Discrimination Loss and Superclass Suppression Loss guide the network’s training,(LPDL)(LSSL)effectively addressing false activation without the need for extra supervision.The CPEWS model proposed in this paper achieves state-of-the-art performance in end-to-end weakly supervised semantic segmentation without additional supervision.The validation set and test set Mean Intersection over Union(MIoU)of PASCAL VOC 2012 dataset achieved 69.8%and 72.6%,respectively.Compared with ToCo(pre trained weight ImageNet-1k),MIoU on the test set is 2.1%higher.In addition,MIoU reached 41.4%on the validation set of the MS COCO 2014 dataset.展开更多
With the advent of deep learning,self-driving schemes based on deep learning are becoming more and more popular.Robust perception-action models should learn from data with different scenarios and real behaviors,while ...With the advent of deep learning,self-driving schemes based on deep learning are becoming more and more popular.Robust perception-action models should learn from data with different scenarios and real behaviors,while current end-to-end model learning is generally limited to training of massive data,innovation of deep network architecture,and learning in-situ model in a simulation environment.Therefore,we introduce a new image style transfer method into data augmentation,and improve the diversity of limited data by changing the texture,contrast ratio and color of the image,and then it is extended to the scenarios that the model has been unobserved before.Inspired by rapid style transfer and artistic style neural algorithms,we propose an arbitrary style generation network architecture,including style transfer network,style learning network,style loss network and multivariate Gaussian distribution function.The style embedding vector is randomly sampled from the multivariate Gaussian distribution and linearly interpolated with the embedded vector predicted by the input image on the style learning network,which provides a set of normalization constants for the style transfer network,and finally realizes the diversity of the image style.In order to verify the effectiveness of the method,image classification and simulation experiments were performed separately.Finally,we built a small-sized smart car experiment platform,and apply the data augmentation technology based on image style transfer drive to the experiment of automatic driving for the first time.The experimental results show that:(1)The proposed scheme can improve the prediction accuracy of the end-to-end model and reduce the model’s error accumulation;(2)the method based on image style transfer provides a new scheme for data augmentation technology,and also provides a solution for the high cost that many deep models rely heavily on a large number of label data.展开更多
End-to-end repair under no or low tension leads to improved outcomes for transected nerves with short gaps,compared to repairs with a graft.However,grafts are typically used to enable a tension-free repair for moderat...End-to-end repair under no or low tension leads to improved outcomes for transected nerves with short gaps,compared to repairs with a graft.However,grafts are typically used to enable a tension-free repair for moderate to large gaps,as excessive tension can cause repairs to fail and catastrophically impede recovery.In this study,we tested the hypothesis that unloading the repair interface by redistributing tension away from the site of repair is a safe and feasible strategy for end-to-end repair of larger nerve gaps.Further,we tested the hypothesis that such an approach does not adversely affect structural and functional regeneration.In this study,we used a rat sciatic nerve injury model to compare the integrity of repair and several regenerative outcomes following end-to-end repairs of nerve gaps of increasing size.In addition,we proposed the use of a novel implantable device to safely repair end-to-end repair of larger nerve gaps by redistributing tension away from the repair interface.Our data suggest that redistriubution of tension away from the site of repair enables safe end-to-end repair of larger gap sizes.In addition,structural and functional measures of regeneration were equal or enhanced in nerves repaired under tension – with or without a tension redistribution device – compared to tension-free repairs.Provided that repair integrity is maintained,end-to-end repairs under tension should be considered as a reasonable surgical strategy.All animal experiments were performed under the approval of the Institutional Animal Care and Use Committee of University of California,San Diego(Protocol S11274).展开更多
Capacity reduction is a major problem faced by wireless mesh networks. An efficient way to alleviate this problem is proper channel assignment. Current end-toend channel assignment schemes usually focus on the case wh...Capacity reduction is a major problem faced by wireless mesh networks. An efficient way to alleviate this problem is proper channel assignment. Current end-toend channel assignment schemes usually focus on the case where channels in distinct frequency bands are assigned to mesh access and backbone, but actually backbone network and access network can use the same IEEE 802.11 technology. Besides, these channel assignment schemes only utilize orthogonal channels to perform channel assignment, and the resulting network interference dramatically degrades network performance. Moreover, Internet-oriented traffic is considered only, and peerto-peer traffic is omitted, or vice versa. The traffic type does not match the practical network. In this paper, we explore how to exploit partially overlapped channels to perform endto-end channel assignment in order to achieve effective end-to-end flow transmissions. The proposed flow-based end-to-end channel assignment schemes can conquer the limitations aforementioned. Simulations reveal that loadaware channel assignment can be applied to networks with stable traffic load, and it can achieve near-optimal performance; Traffic-irrelevant channel assignment is suitable for networks with frequent change of traffic load,and it can achieve good balance between performance and overhead. Also, partially overlapped channels' capability of improving network performance is situation-dependent, they should be used carefully.展开更多
Traditional speech recognition model based on deep neural network(DNN)and hidden Markov model(HMM)is a complex and multi-module system.In other words,optimization goals may differ between modules in traditional model....Traditional speech recognition model based on deep neural network(DNN)and hidden Markov model(HMM)is a complex and multi-module system.In other words,optimization goals may differ between modules in traditional model.Besides,additional language resources are required,such as pronunciation dictionary and language model.To eliminate the drawbacks of traditional model,we hereby propose an end-to-end speech recognition method,where connectionist temporal classification(CTC)and attention are integrated for decoding.In our model,the complex modules are replaced by a single deep network.Our model mainly consists of encoder and decoder.The encoder is constructed by bidirectional long short-term memory(BLSTM)with a triangular structure for feature extraction.The decoder based on CTC-attention decoding utilizes advanced features extracted by shared encoder for training and decoding.The experimental results on the Vox Forge dataset indicate that end-to-end method is superior to basic CTC and attention-based encoder-decoder decoding,and the character error rate(CER)is reduced to 12.9%without using any language model.展开更多
End-to-end delay is one of the most important characteristics of Internet end-to-end packet dynamics, which can be applied to quality of services （OoS） management, service level agreement （SLA） management, congest...End-to-end delay is one of the most important characteristics of Internet end-to-end packet dynamics, which can be applied to quality of services （OoS） management, service level agreement （SLA） management, congestion control algorithm development, etc. Nonstationarity and nonlinearity are found by the analysis of various delay series measured from different links. The fact that different types of links have different degree of Self-Similarity is also obtained. By constructing appropriate network architecture and neural functions, functional networks can be used to model the Internet end-to-end nonlinear delay time series. Furthermore, by using adaptive parameter studying algorithm, the nonstationarity can also be well modeled. The numerical results show that the provided functional network architecture and adaptive algorithm can precisely characterize the Internet end-to-end delay dynamics.展开更多
Environment perception is one of the most critical technology of intelligent transportation systems(ITS).Motion interaction between multiple vehicles in ITS makes it important to perform multi-object tracking(MOT).How...Environment perception is one of the most critical technology of intelligent transportation systems(ITS).Motion interaction between multiple vehicles in ITS makes it important to perform multi-object tracking(MOT).However,most existing MOT algorithms follow the tracking-by-detection framework,which separates detection and tracking into two independent segments and limit the global efciency.Recently,a few algorithms have combined feature extraction into one network;however,the tracking portion continues to rely on data association,and requires com‑plex post-processing for life cycle management.Those methods do not combine detection and tracking efciently.This paper presents a novel network to realize joint multi-object detection and tracking in an end-to-end manner for ITS,named as global correlation network(GCNet).Unlike most object detection methods,GCNet introduces a global correlation layer for regression of absolute size and coordinates of bounding boxes,instead of ofsetting predictions.The pipeline of detection and tracking in GCNet is conceptually simple,and does not require compli‑cated tracking strategies such as non-maximum suppression and data association.GCNet was evaluated on a multivehicle tracking dataset,UA-DETRAC,demonstrating promising performance compared to state-of-the-art detectors and trackers.展开更多
Person search mainly consists of two submissions,namely Person Detection and Person Re-identification(reID).Existing approaches are primarily based on Faster R-CNN and Convolutional Neural Network(CNN)(e.g.,ResNet).Wh...Person search mainly consists of two submissions,namely Person Detection and Person Re-identification(reID).Existing approaches are primarily based on Faster R-CNN and Convolutional Neural Network(CNN)(e.g.,ResNet).While these structures may detect high-quality bounding boxes,they seem to degrade the performance of re-ID.To address this issue,this paper proposes a Dual-Transformer Head Network(DTHN)for end-to-end person search,which contains two independent Transformer heads,a box head for detecting the bounding box and extracting efficient bounding box feature,and a re-ID head for capturing high-quality re-ID features for the re-ID task.Specifically,after the image goes through the ResNet backbone network to extract features,the Region Proposal Network(RPN)proposes possible bounding boxes.The box head then extracts more efficient features within these bounding boxes for detection.Following this,the re-ID head computes the occluded attention of the features in these bounding boxes and distinguishes them from other persons or backgrounds.Extensive experiments on two widely used benchmark datasets,CUHK-SYSU and PRW,achieve state-of-the-art performance levels,94.9 mAP and 95.3 top-1 scores on the CUHK-SYSU dataset,and 51.6 mAP and 87.6 top-1 scores on the PRW dataset,which demonstrates the advantages of this paper’s approach.The efficiency comparison also shows our method is highly efficient in both time and space.展开更多
With the emergence of large-scale knowledge base,how to use triple information to generate natural questions is a key technology in question answering systems.The traditional way of generating questions require a lot ...With the emergence of large-scale knowledge base,how to use triple information to generate natural questions is a key technology in question answering systems.The traditional way of generating questions require a lot of manual intervention and produce lots of noise.To solve these problems,we propose a joint model based on semi-automated model and End-to-End neural network to automatically generate questions.The semi-automated model can generate question templates and real questions combining the knowledge base and center graph.The End-to-End neural network directly sends the knowledge base and real questions to BiLSTM network.Meanwhile,the attention mechanism is utilized in the decoding layer,which makes the triples and generated questions more relevant.Finally,the experimental results on SimpleQuestions demonstrate the effectiveness of the proposed approach.展开更多
The end-to-end separation algorithm with superior performance in the field of speech separation has not been effectively used in music separation.Moreover,since music signals are often dual channel data with a high sa...The end-to-end separation algorithm with superior performance in the field of speech separation has not been effectively used in music separation.Moreover,since music signals are often dual channel data with a high sampling rate,how to model longsequence data and make rational use of the relevant information between channels is also an urgent problem to be solved.In order to solve the above problems,the performance of the end-to-end music separation algorithm is enhanced by improving the network structure.Our main contributions include the following:(1)A more reasonable densely connected U-Net is designed to capture the long-term characteristics of music,such as main melody,tone and so on.(2)On this basis,the multi-head attention and dualpath transformer are introduced in the separation module.Channel attention units are applied recursively on the feature map of each layer of the network,enabling the network to perform long-sequence separation.Experimental results show that after the introduction of the channel attention,the performance of the proposed algorithm has a stable improvement compared with the baseline system.On the MUSDB18 dataset,the average score of the separated audio exceeds that of the current best-performing music separation algorithm based on the time-frequency domain(T-F domain).展开更多
This paper presents the multi-step Q-learning(MQL)algorithm as an autonomic approach to thejoint radio resource management(JRRM)among heterogeneous radio access technologies(RATs)in theB3G environment.Through the'...This paper presents the multi-step Q-learning(MQL)algorithm as an autonomic approach to thejoint radio resource management(JRRM)among heterogeneous radio access technologies(RATs)in theB3G environment.Through the'trial-and-error'on-line learning process,the JRRM controller can con-verge to the optimized admission control policy.The JRRM controller learns to give the best allocation foreach session in terms of both the access RAT and the service bandwidth.Simulation results show that theproposed algorithm realizes the autonomy of JRRM and achieves well trade-off between the spectrum utilityand the blocking probability comparing to the load-balancing algorithm and the utility-maximizing algo-rithm.Besides,the proposed algorithm has better online performances and convergence speed than theone-step Q-learning(QL)algorithm.Therefore,the user statisfaction degree could be improved also.展开更多
An adaptive joint source channel bit allocation method for video communications over error-prone channel is proposed.To protect the bit-streams from the channel bit errors,the rate compatible punctured convolution(RCP...An adaptive joint source channel bit allocation method for video communications over error-prone channel is proposed.To protect the bit-streams from the channel bit errors,the rate compatible punctured convolution(RCPC)code is used to produce coding rates varying from 4/5 to 1/2 using the same encoder and the Viterbi decoder.An expected end-to-end distortion model was presented to estimate the distortion introduced in compressed source coding due to quantization and channel bit errors jointly.Based on the proposed end-to-end distortion model,an adaptive joint source-channel bit allocation method was proposed under time-varying error-prone channel conditions.Simulated results show that the proposed methods could utilize the available channel capacity more efficiently and achieve better video quality than the other fixed coding-based bit allocation methods when transmitting over error-prone channels.展开更多
We proposed a method using latent regression Bayesian network (LRBN) toextract the shared speech feature for the input of end-to-end speech recognition model.The structure of LRBN is compact and its parameter learning...We proposed a method using latent regression Bayesian network (LRBN) toextract the shared speech feature for the input of end-to-end speech recognition model.The structure of LRBN is compact and its parameter learning is fast. Compared withConvolutional Neural Network, it has a simpler and understood structure and lessparameters to learn. Experimental results show that the advantage of hybridLRBN/Bidirectional Long Short-Term Memory-Connectionist Temporal Classificationarchitecture for Tibetan multi-dialect speech recognition, and demonstrate the LRBN ishelpful to differentiate among multiple language speech sets.展开更多
The ubiquity of instant messaging services on mobile devices and their use of end-to-end encryption in safeguarding the privacy of their users have become a concern for some governments. WhatsApp messaging service has...The ubiquity of instant messaging services on mobile devices and their use of end-to-end encryption in safeguarding the privacy of their users have become a concern for some governments. WhatsApp messaging service has emerged as the most popular messaging app on mobile devices today. It uses end-to-end encryption which makes government and secret services efforts to combat organized crime, terrorists, and child pornographers technically impossible. Governments would like a “backdoor” into such apps, to use in accessing messages and have emphasized that they will only use the “backdoor” if there is a credible threat to national security. Users of WhatsApp have however, argued against a “backdoor”;they claim a “backdoor” would not only be an infringement of their privacy, but that hackers could also take advantage of it. In light of this security and privacy conflict between the end users of WhatsApp and government’s need to access messages in order to thwart potential terror attacks, this paper presents the advantages of maintaining E2EE in WhatsApp and why governments should not be allowed a “backdoor” to access users’ messages. This research presents the benefits encryption has on consumer security and privacy, and also on the challenges it poses to public safety and national security.展开更多
The concept of semantic communication provides a novel approach for applications in scenarios with limited communication resources.In this paper,we propose an end-to-end(E2E)semantic molecular communication system,aim...The concept of semantic communication provides a novel approach for applications in scenarios with limited communication resources.In this paper,we propose an end-to-end(E2E)semantic molecular communication system,aiming to enhance the efficiency of molecular communication systems by reducing the transmitted information.Specifically,following the joint source channel coding paradigm,the network is designed to encode the task-relevant information into the concentration of the information molecules,which is robust to the degradation of the molecular communication channel.Furthermore,we propose a channel network to enable the E2E learning over the non-differentiable molecular channel.Experimental results demonstrate the superior performance of the semantic molecular communication system over the conventional methods in classification tasks.展开更多
Blockchain is an emerging decentralized technology of electronic voting.The current main consensus protocols are not flexible enough to manage the distributed blockchain nodes to achieve high efficiency of consensus.F...Blockchain is an emerging decentralized technology of electronic voting.The current main consensus protocols are not flexible enough to manage the distributed blockchain nodes to achieve high efficiency of consensus.For practical implementation,the consensus based on random linear block code(RLBC)is proposed and applied to blockchain voting scheme.Along with achieving the record correctness and consistency among all nodes,the consensus method indicates the active and inactive consensus nodes.This ability can assist the management of consensus nodes and restrain the generating of chain forks.To achieve end-to-end verifiability,cast-or-audit and randomized partial checking(RPC)are used in the proposed scheme.The voter can verify the high probability of correctness in ballot encryption and decryption.The experiments illustrate that the efficiency of proposed consensus is suitable for blockchain.The proposed electronic voting scheme is adapted to practical implementation of voting.展开更多
Interdisciplinary applications between information technology and geriatrics have been accelerated in recent years by the advancement of artificial intelligence,cloud computing,and 5G technology,among others.Meanwhile...Interdisciplinary applications between information technology and geriatrics have been accelerated in recent years by the advancement of artificial intelligence,cloud computing,and 5G technology,among others.Meanwhile,applications developed by using the above technologies make it possible to predict the risk of age-related diseases early,which can give caregivers time to intervene and reduce the risk,potentially improving the health span of the elderly.However,the popularity of these applications is still limited for several reasons.For example,many older people are unable or unwilling to use mobile applications or devices(e.g.smartphones)because they are relatively complex operations or time-consuming for older people.In this work,we design and implement an end-to-end framework and integrate it with the WeChat platform to make it easily accessible to elders.In this work,multifactorial geriatric assessment data can be collected.Then,stacked machine learning models are trained to assess and predict the incidence of common diseases in the elderly.Experimental results show that our framework can not only provide more accurate prediction(precision:0.8713,recall:0.8212)for several common elderly diseases,but also very low timeconsuming(28.6 s)within a workflow compared to some existing similar applications.展开更多
基金supported by the National Natural Science Foundation of China(62106214)the Hebei Natural Science Foundation(D2024203008)the Provincial Key Laboratory Performance Subsidy Project(22567612H).
文摘In recent years,audio pattern recognition has emerged as a key area of research,driven by its applications in human-computer interaction,robotics,and healthcare.Traditional methods,which rely heavily on handcrafted features such asMel filters,often suffer frominformation loss and limited feature representation capabilities.To address these limitations,this study proposes an innovative end-to-end audio pattern recognition framework that directly processes raw audio signals,preserving original information and extracting effective classification features.The proposed framework utilizes a dual-branch architecture:a global refinement module that retains channel and temporal details and a multi-scale embedding module that captures high-level semantic information.Additionally,a guided fusion module integrates complementary features from both branches,ensuring a comprehensive representation of audio data.Specifically,the multi-scale audio context embedding module is designed to effectively extract spatiotemporal dependencies,while the global refinement module aggregates multi-scale channel and temporal cues for enhanced modeling.The guided fusion module leverages these features to achieve efficient integration of complementary information,resulting in improved classification accuracy.Experimental results demonstrate the model’s superior performance on multiple datasets,including ESC-50,UrbanSound8K,RAVDESS,and CREMA-D,with classification accuracies of 93.25%,90.91%,92.36%,and 70.50%,respectively.These results highlight the robustness and effectiveness of the proposed framework,which significantly outperforms existing approaches.By addressing critical challenges such as information loss and limited feature representation,thiswork provides newinsights and methodologies for advancing audio classification and multimodal interaction systems.
文摘Machine-to-machine (M2M) communication networks consist of resource-constrained autonomous devices, also known as autonomous Internet of things (IoTs) or machine-type communication devices (MTCDs) which act as a backbone for Industrial IoT, smart cities, and other autonomous systems. Due to the limited computing and memory capacity, these devices cannot maintain strong security if conventional security methods are applied such as heavy encryption. This article proposed a novel lightweight mutual authentication scheme including elliptic curve cryptography (ECC) driven end-to-end encryption through curve25519 such as (i): efficient end-to-end encrypted communication with pre-calculation strategy using curve25519;and (ii): elliptic curve Diffie-Hellman (ECDH) based mutual authentication technique through a novel lightweight hash function. The proposed scheme attempts to efficiently counter all known perception layer security threats. Moreover, the pre-calculated key generation strategy resulted in cost-effective encryption with 192-bit curve security. It showed comparative efficiency in key strength, and curve strength compared with similar authentication schemes in terms of computational and memory cost, communication performance and encryption robustness.
基金supported by the National Natural Science Foundation of China(Grant Nos.62227821,62025503,and 62205199).
文摘To capture the nonlinear dynamics and gain evolution in chirped pulse amplification(CPA)systems,the split-step Fourier method and the fourth-order Runge–Kutta method are integrated to iteratively address the generalized nonlinear Schrödinger equation and the rate equations.However,this approach is burdened by substantial computational demands,resulting in significant time expenditures.In the context of intelligent laser optimization and inverse design,the necessity for numerous simulations further exacerbates this issue,highlighting the need for fast and accurate simulation methodologies.Here,we introduce an end-to-end model augmented with active learning(E2E-AL)with decent generalization through different dedicated embedding methods over various parameters.On an identical computational platform,the artificial intelligence–driven model is 2000 times faster than the conventional simulation method.Benefiting from the active learning strategy,the E2E-AL model achieves decent precision with only two-thirds of the training samples compared with the case without such a strategy.Furthermore,we demonstrate a multi-objective inverse design of the CPA systems enabled by the E2E-AL model.The E2E-AL framework manifests the potential of becoming a standard approach for the rapid and accurate modeling of ultrafast lasers and is readily extended to simulate other complex systems.
基金funding from the following sources:National Natural Science Foundation of China(U1904119)Research Programs of Henan Science and Technology Department(232102210054)+3 种基金Chongqing Natural Science Foundation(CSTB2023NSCQ-MSX0070)Henan Province Key Research and Development Project(231111212000)Aviation Science Foundation(20230001055002)supported by Henan Center for Outstanding Overseas Scientists(GZS2022011).
文摘The primary challenge in weakly supervised semantic segmentation is effectively leveraging weak annotations while minimizing the performance gap compared to fully supervised methods.End-to-end model designs have gained significant attention for improving training efficiency.Most current algorithms rely on Convolutional Neural Networks(CNNs)for feature extraction.Although CNNs are proficient at capturing local features,they often struggle with global context,leading to incomplete and false Class Activation Mapping(CAM).To address these limitations,this work proposes a Contextual Prototype-Based End-to-End Weakly Supervised Semantic Segmentation(CPEWS)model,which improves feature extraction by utilizing the Vision Transformer(ViT).By incorporating its intermediate feature layers to preserve semantic information,this work introduces the Intermediate Supervised Module(ISM)to supervise the final layer’s output,reducing boundary ambiguity and mitigating issues related to incomplete activation.Additionally,the Contextual Prototype Module(CPM)generates class-specific prototypes,while the proposed Prototype Discrimination Loss and Superclass Suppression Loss guide the network’s training,(LPDL)(LSSL)effectively addressing false activation without the need for extra supervision.The CPEWS model proposed in this paper achieves state-of-the-art performance in end-to-end weakly supervised semantic segmentation without additional supervision.The validation set and test set Mean Intersection over Union(MIoU)of PASCAL VOC 2012 dataset achieved 69.8%and 72.6%,respectively.Compared with ToCo(pre trained weight ImageNet-1k),MIoU on the test set is 2.1%higher.In addition,MIoU reached 41.4%on the validation set of the MS COCO 2014 dataset.
基金the National Natural Science Foundation of China(51965008)Science and Technology projects of Guizhou[2018]2168Excellent Young Researcher Project of Guizhou[2017]5630.
文摘With the advent of deep learning,self-driving schemes based on deep learning are becoming more and more popular.Robust perception-action models should learn from data with different scenarios and real behaviors,while current end-to-end model learning is generally limited to training of massive data,innovation of deep network architecture,and learning in-situ model in a simulation environment.Therefore,we introduce a new image style transfer method into data augmentation,and improve the diversity of limited data by changing the texture,contrast ratio and color of the image,and then it is extended to the scenarios that the model has been unobserved before.Inspired by rapid style transfer and artistic style neural algorithms,we propose an arbitrary style generation network architecture,including style transfer network,style learning network,style loss network and multivariate Gaussian distribution function.The style embedding vector is randomly sampled from the multivariate Gaussian distribution and linearly interpolated with the embedded vector predicted by the input image on the style learning network,which provides a set of normalization constants for the style transfer network,and finally realizes the diversity of the image style.In order to verify the effectiveness of the method,image classification and simulation experiments were performed separately.Finally,we built a small-sized smart car experiment platform,and apply the data augmentation technology based on image style transfer drive to the experiment of automatic driving for the first time.The experimental results show that:(1)The proposed scheme can improve the prediction accuracy of the end-to-end model and reduce the model’s error accumulation;(2)the method based on image style transfer provides a new scheme for data augmentation technology,and also provides a solution for the high cost that many deep models rely heavily on a large number of label data.
基金supported by the Department of Veterans Affairs(VA MERIT IRX001471A to SBS)
文摘End-to-end repair under no or low tension leads to improved outcomes for transected nerves with short gaps,compared to repairs with a graft.However,grafts are typically used to enable a tension-free repair for moderate to large gaps,as excessive tension can cause repairs to fail and catastrophically impede recovery.In this study,we tested the hypothesis that unloading the repair interface by redistributing tension away from the site of repair is a safe and feasible strategy for end-to-end repair of larger nerve gaps.Further,we tested the hypothesis that such an approach does not adversely affect structural and functional regeneration.In this study,we used a rat sciatic nerve injury model to compare the integrity of repair and several regenerative outcomes following end-to-end repairs of nerve gaps of increasing size.In addition,we proposed the use of a novel implantable device to safely repair end-to-end repair of larger nerve gaps by redistributing tension away from the repair interface.Our data suggest that redistriubution of tension away from the site of repair enables safe end-to-end repair of larger gap sizes.In addition,structural and functional measures of regeneration were equal or enhanced in nerves repaired under tension – with or without a tension redistribution device – compared to tension-free repairs.Provided that repair integrity is maintained,end-to-end repairs under tension should be considered as a reasonable surgical strategy.All animal experiments were performed under the approval of the Institutional Animal Care and Use Committee of University of California,San Diego(Protocol S11274).
基金supported by the National Natural Science Foundation of China under Grants No.61373124
文摘Capacity reduction is a major problem faced by wireless mesh networks. An efficient way to alleviate this problem is proper channel assignment. Current end-toend channel assignment schemes usually focus on the case where channels in distinct frequency bands are assigned to mesh access and backbone, but actually backbone network and access network can use the same IEEE 802.11 technology. Besides, these channel assignment schemes only utilize orthogonal channels to perform channel assignment, and the resulting network interference dramatically degrades network performance. Moreover, Internet-oriented traffic is considered only, and peerto-peer traffic is omitted, or vice versa. The traffic type does not match the practical network. In this paper, we explore how to exploit partially overlapped channels to perform endto-end channel assignment in order to achieve effective end-to-end flow transmissions. The proposed flow-based end-to-end channel assignment schemes can conquer the limitations aforementioned. Simulations reveal that loadaware channel assignment can be applied to networks with stable traffic load, and it can achieve near-optimal performance; Traffic-irrelevant channel assignment is suitable for networks with frequent change of traffic load,and it can achieve good balance between performance and overhead. Also, partially overlapped channels' capability of improving network performance is situation-dependent, they should be used carefully.
文摘Traditional speech recognition model based on deep neural network(DNN)and hidden Markov model(HMM)is a complex and multi-module system.In other words,optimization goals may differ between modules in traditional model.Besides,additional language resources are required,such as pronunciation dictionary and language model.To eliminate the drawbacks of traditional model,we hereby propose an end-to-end speech recognition method,where connectionist temporal classification(CTC)and attention are integrated for decoding.In our model,the complex modules are replaced by a single deep network.Our model mainly consists of encoder and decoder.The encoder is constructed by bidirectional long short-term memory(BLSTM)with a triangular structure for feature extraction.The decoder based on CTC-attention decoding utilizes advanced features extracted by shared encoder for training and decoding.The experimental results on the Vox Forge dataset indicate that end-to-end method is superior to basic CTC and attention-based encoder-decoder decoding,and the character error rate(CER)is reduced to 12.9%without using any language model.
基金This project was supported by the National Natural Science Foundation of China (60132030 60572147)
文摘End-to-end delay is one of the most important characteristics of Internet end-to-end packet dynamics, which can be applied to quality of services （OoS） management, service level agreement （SLA） management, congestion control algorithm development, etc. Nonstationarity and nonlinearity are found by the analysis of various delay series measured from different links. The fact that different types of links have different degree of Self-Similarity is also obtained. By constructing appropriate network architecture and neural functions, functional networks can be used to model the Internet end-to-end nonlinear delay time series. Furthermore, by using adaptive parameter studying algorithm, the nonstationarity can also be well modeled. The numerical results show that the provided functional network architecture and adaptive algorithm can precisely characterize the Internet end-to-end delay dynamics.
基金Supported by National Key Research and Development Program of China(Grant No.2021YFB1600402)National Natural Science Foundation of China(Grant No.52072212)+1 种基金Dongfeng USharing Technology Co.,Ltd.,China Intelli‑gent and Connected Vehicles(Beijing)Research Institute Co.,Ltd.“Shuimu Tsinghua Scholarship”of Tsinghua University of China.
文摘Environment perception is one of the most critical technology of intelligent transportation systems(ITS).Motion interaction between multiple vehicles in ITS makes it important to perform multi-object tracking(MOT).However,most existing MOT algorithms follow the tracking-by-detection framework,which separates detection and tracking into two independent segments and limit the global efciency.Recently,a few algorithms have combined feature extraction into one network;however,the tracking portion continues to rely on data association,and requires com‑plex post-processing for life cycle management.Those methods do not combine detection and tracking efciently.This paper presents a novel network to realize joint multi-object detection and tracking in an end-to-end manner for ITS,named as global correlation network(GCNet).Unlike most object detection methods,GCNet introduces a global correlation layer for regression of absolute size and coordinates of bounding boxes,instead of ofsetting predictions.The pipeline of detection and tracking in GCNet is conceptually simple,and does not require compli‑cated tracking strategies such as non-maximum suppression and data association.GCNet was evaluated on a multivehicle tracking dataset,UA-DETRAC,demonstrating promising performance compared to state-of-the-art detectors and trackers.
基金supported by the Natural Science Foundation of Shanghai under Grant 21ZR1426500the National Natural Science Foundation of China under Grant 61873160.
文摘Person search mainly consists of two submissions,namely Person Detection and Person Re-identification(reID).Existing approaches are primarily based on Faster R-CNN and Convolutional Neural Network(CNN)(e.g.,ResNet).While these structures may detect high-quality bounding boxes,they seem to degrade the performance of re-ID.To address this issue,this paper proposes a Dual-Transformer Head Network(DTHN)for end-to-end person search,which contains two independent Transformer heads,a box head for detecting the bounding box and extracting efficient bounding box feature,and a re-ID head for capturing high-quality re-ID features for the re-ID task.Specifically,after the image goes through the ResNet backbone network to extract features,the Region Proposal Network(RPN)proposes possible bounding boxes.The box head then extracts more efficient features within these bounding boxes for detection.Following this,the re-ID head computes the occluded attention of the features in these bounding boxes and distinguishes them from other persons or backgrounds.Extensive experiments on two widely used benchmark datasets,CUHK-SYSU and PRW,achieve state-of-the-art performance levels,94.9 mAP and 95.3 top-1 scores on the CUHK-SYSU dataset,and 51.6 mAP and 87.6 top-1 scores on the PRW dataset,which demonstrates the advantages of this paper’s approach.The efficiency comparison also shows our method is highly efficient in both time and space.
基金supported by National Nature Science Foundation(No.61501529,No.61331013)National Language Committee Project of China(No.ZDI125-36)Young Teachers'Scientific Research Project in Minzu University of China.
文摘With the emergence of large-scale knowledge base,how to use triple information to generate natural questions is a key technology in question answering systems.The traditional way of generating questions require a lot of manual intervention and produce lots of noise.To solve these problems,we propose a joint model based on semi-automated model and End-to-End neural network to automatically generate questions.The semi-automated model can generate question templates and real questions combining the knowledge base and center graph.The End-to-End neural network directly sends the knowledge base and real questions to BiLSTM network.Meanwhile,the attention mechanism is utilized in the decoding layer,which makes the triples and generated questions more relevant.Finally,the experimental results on SimpleQuestions demonstrate the effectiveness of the proposed approach.
基金National Natural Science Foundation of China,Grant/Award Number:62071039Beijing Natural Science Foundation,Grant/Award Number:L223033。
文摘The end-to-end separation algorithm with superior performance in the field of speech separation has not been effectively used in music separation.Moreover,since music signals are often dual channel data with a high sampling rate,how to model longsequence data and make rational use of the relevant information between channels is also an urgent problem to be solved.In order to solve the above problems,the performance of the end-to-end music separation algorithm is enhanced by improving the network structure.Our main contributions include the following:(1)A more reasonable densely connected U-Net is designed to capture the long-term characteristics of music,such as main melody,tone and so on.(2)On this basis,the multi-head attention and dualpath transformer are introduced in the separation module.Channel attention units are applied recursively on the feature map of each layer of the network,enabling the network to perform long-sequence separation.Experimental results show that after the introduction of the channel attention,the performance of the proposed algorithm has a stable improvement compared with the baseline system.On the MUSDB18 dataset,the average score of the separated audio exceeds that of the current best-performing music separation algorithm based on the time-frequency domain(T-F domain).
基金the National Natural Science Foundation of China(No.60632030)the National High Technology Research and Development Program of China(No.2006AA01Z276)
文摘This paper presents the multi-step Q-learning(MQL)algorithm as an autonomic approach to thejoint radio resource management(JRRM)among heterogeneous radio access technologies(RATs)in theB3G environment.Through the'trial-and-error'on-line learning process,the JRRM controller can con-verge to the optimized admission control policy.The JRRM controller learns to give the best allocation foreach session in terms of both the access RAT and the service bandwidth.Simulation results show that theproposed algorithm realizes the autonomy of JRRM and achieves well trade-off between the spectrum utilityand the blocking probability comparing to the load-balancing algorithm and the utility-maximizing algo-rithm.Besides,the proposed algorithm has better online performances and convergence speed than theone-step Q-learning(QL)algorithm.Therefore,the user statisfaction degree could be improved also.
基金National High-Tech Research and Development Plan of China(No.2003AA1Z2130)Science and Technology Project of Zhejiang Province,China(No.2006C11200)
文摘An adaptive joint source channel bit allocation method for video communications over error-prone channel is proposed.To protect the bit-streams from the channel bit errors,the rate compatible punctured convolution(RCPC)code is used to produce coding rates varying from 4/5 to 1/2 using the same encoder and the Viterbi decoder.An expected end-to-end distortion model was presented to estimate the distortion introduced in compressed source coding due to quantization and channel bit errors jointly.Based on the proposed end-to-end distortion model,an adaptive joint source-channel bit allocation method was proposed under time-varying error-prone channel conditions.Simulated results show that the proposed methods could utilize the available channel capacity more efficiently and achieve better video quality than the other fixed coding-based bit allocation methods when transmitting over error-prone channels.
文摘We proposed a method using latent regression Bayesian network (LRBN) toextract the shared speech feature for the input of end-to-end speech recognition model.The structure of LRBN is compact and its parameter learning is fast. Compared withConvolutional Neural Network, it has a simpler and understood structure and lessparameters to learn. Experimental results show that the advantage of hybridLRBN/Bidirectional Long Short-Term Memory-Connectionist Temporal Classificationarchitecture for Tibetan multi-dialect speech recognition, and demonstrate the LRBN ishelpful to differentiate among multiple language speech sets.
文摘The ubiquity of instant messaging services on mobile devices and their use of end-to-end encryption in safeguarding the privacy of their users have become a concern for some governments. WhatsApp messaging service has emerged as the most popular messaging app on mobile devices today. It uses end-to-end encryption which makes government and secret services efforts to combat organized crime, terrorists, and child pornographers technically impossible. Governments would like a “backdoor” into such apps, to use in accessing messages and have emphasized that they will only use the “backdoor” if there is a credible threat to national security. Users of WhatsApp have however, argued against a “backdoor”;they claim a “backdoor” would not only be an infringement of their privacy, but that hackers could also take advantage of it. In light of this security and privacy conflict between the end users of WhatsApp and government’s need to access messages in order to thwart potential terror attacks, this paper presents the advantages of maintaining E2EE in WhatsApp and why governments should not be allowed a “backdoor” to access users’ messages. This research presents the benefits encryption has on consumer security and privacy, and also on the challenges it poses to public safety and national security.
基金supported by the Beijing Natural Science Foundation(L211012)the Natural Science Foundation of China(62122012,62221001)the Fundamental Research Funds for the Central Universities(2022JBQY004)。
文摘The concept of semantic communication provides a novel approach for applications in scenarios with limited communication resources.In this paper,we propose an end-to-end(E2E)semantic molecular communication system,aiming to enhance the efficiency of molecular communication systems by reducing the transmitted information.Specifically,following the joint source channel coding paradigm,the network is designed to encode the task-relevant information into the concentration of the information molecules,which is robust to the degradation of the molecular communication channel.Furthermore,we propose a channel network to enable the E2E learning over the non-differentiable molecular channel.Experimental results demonstrate the superior performance of the semantic molecular communication system over the conventional methods in classification tasks.
基金Supported by the National Natural Science Foundation of China(No.61501064)Sichuan Technology Support Program(No.2015GZ0088)Guangxi Key Laboratory of Hybrid Computation and IC Design Analysis(No.HCIC201502,HCIC201701)。
文摘Blockchain is an emerging decentralized technology of electronic voting.The current main consensus protocols are not flexible enough to manage the distributed blockchain nodes to achieve high efficiency of consensus.For practical implementation,the consensus based on random linear block code(RLBC)is proposed and applied to blockchain voting scheme.Along with achieving the record correctness and consistency among all nodes,the consensus method indicates the active and inactive consensus nodes.This ability can assist the management of consensus nodes and restrain the generating of chain forks.To achieve end-to-end verifiability,cast-or-audit and randomized partial checking(RPC)are used in the proposed scheme.The voter can verify the high probability of correctness in ballot encryption and decryption.The experiments illustrate that the efficiency of proposed consensus is suitable for blockchain.The proposed electronic voting scheme is adapted to practical implementation of voting.
基金supported by Xi’an University of Finance and Economics Scientific Research Support Program(No.21FCZD03)Shaanxi Education Department Research Program(No.22JK0077)National Statistical Science Research Project(Nos.2021LZ40,2022LZ38)。
文摘Interdisciplinary applications between information technology and geriatrics have been accelerated in recent years by the advancement of artificial intelligence,cloud computing,and 5G technology,among others.Meanwhile,applications developed by using the above technologies make it possible to predict the risk of age-related diseases early,which can give caregivers time to intervene and reduce the risk,potentially improving the health span of the elderly.However,the popularity of these applications is still limited for several reasons.For example,many older people are unable or unwilling to use mobile applications or devices(e.g.smartphones)because they are relatively complex operations or time-consuming for older people.In this work,we design and implement an end-to-end framework and integrate it with the WeChat platform to make it easily accessible to elders.In this work,multifactorial geriatric assessment data can be collected.Then,stacked machine learning models are trained to assess and predict the incidence of common diseases in the elderly.Experimental results show that our framework can not only provide more accurate prediction(precision:0.8713,recall:0.8212)for several common elderly diseases,but also very low timeconsuming(28.6 s)within a workflow compared to some existing similar applications.
