Robotic computing systems play an important role in enabling intelligent robotic tasks through intelligent algo-rithms and supporting hardware.In recent years,the evolution of robotic algorithms indicates a roadmap fr...Robotic computing systems play an important role in enabling intelligent robotic tasks through intelligent algo-rithms and supporting hardware.In recent years,the evolution of robotic algorithms indicates a roadmap from traditional robotics to hierarchical and end-to-end models.This algorithmic advancement poses a critical challenge in achieving balanced system-wide performance.Therefore,algorithm-hardware co-design has emerged as the primary methodology,which ana-lyzes algorithm behaviors on hardware to identify common computational properties.These properties can motivate algo-rithm optimization to reduce computational complexity and hardware innovation from architecture to circuit for high performance and high energy efficiency.We then reviewed recent works on robotic and embodied AI algorithms and computing hard-ware to demonstrate this algorithm-hardware co-design methodology.In the end,we discuss future research opportunities by answering two questions:(1)how to adapt the computing platforms to the rapid evolution of embodied AI algorithms,and(2)how to transform the potential of emerging hardware innovations into end-to-end inference improvements.展开更多
Compared to the last decade when the convolution neu-ral network(CNN)dominated the research field,machine learn-ing(ML)algorithms have reached a pivotal moment called the generative artificial intelligence(AI)era.With...Compared to the last decade when the convolution neu-ral network(CNN)dominated the research field,machine learn-ing(ML)algorithms have reached a pivotal moment called the generative artificial intelligence(AI)era.With the emer-gence of large-scale foundation models[1],such as large multi-modal model(LMM)GPT-4[2]and text-to-image generative model DALL·E[3].展开更多
植物图片包含植物生境、物种组成、形态特征、物候等相关信息,是野外调查和植物记录的重要资料。无人机可按照设定程序定时、定航线拍摄植物,获取植物图片拍摄地的精准位置信息,进而实现周期化的植物拍摄和调查。本图片数据集系于2022–...植物图片包含植物生境、物种组成、形态特征、物候等相关信息,是野外调查和植物记录的重要资料。无人机可按照设定程序定时、定航线拍摄植物,获取植物图片拍摄地的精准位置信息,进而实现周期化的植物拍摄和调查。本图片数据集系于2022–2023年在内蒙古呼伦贝尔湿润草原、锡林浩特典型草原、鄂尔多斯干旱草原选地,依照《草地植物多样性无人机调查技术规范》(T/CSES 123-2023)团体标准,以DJI MINI 3 PRO无人机采集而来,并以人工框选和鉴定为主、目标检测和智能识别模型处理为辅的方式进行了图像中的植物框选和鉴定。本数据集包含了19科32属40种植物的4000幅图片、植物物种名称、植物科属信息、采集时间、采集点海拔、经纬度。本数据集可以为相关草地植物的形态、分布、物候等信息检索以及智能识别模型构建提供数据支撑。展开更多
Being a nonlinear operator,fractional derivatives can affect the enforcement of existence at any given time.As a result,the memory effect has an impact on all nonlinear processes modeled by fractional order differenti...Being a nonlinear operator,fractional derivatives can affect the enforcement of existence at any given time.As a result,the memory effect has an impact on all nonlinear processes modeled by fractional order differential equations(FODEs).The goal of this study is to increase the fractional model of the TB virus’s(FMTBV)accuracy.Stochastic solvers have never been used to solve FMTBV previously.The Bayesian regularized artificial(BRA)method and neural networks(NNs),often referred to as BRA-NNs,were used to solve the FMTBV model.Each scenario features five occurrences that each reflect a different order of derivatives,ranging from 0.8,0.85,0.9,0.95,and 1,as well as five potential rates for different parameters.Training data made up 90%of the data,testing data made up 5%,and validation data made up 5%of the data used to illustrate the FMTBV’s approximations.To verify that the BRA-NNs were correct,the generated simulations were described in the following solutions using the FOLotkaVolterra approach in MATLAB.Comprehensive Simulink results in terms of mean square error,error histogram,and regression analysis investigations further highlight the competence,dependability,and accuracy of the suggested BRA-NNs.展开更多
SARS-CoV-2 infection and vaccination both trigger immune responses. The former leads to naturally acquired immunity, while the latter induces active immunity through artificial means. However, the distinct immune effe...SARS-CoV-2 infection and vaccination both trigger immune responses. The former leads to naturally acquired immunity, while the latter induces active immunity through artificial means. However, the distinct immune effects of vaccination and infection, as well as their underlying mechanisms, require further clarification. In this study, we compared the peripheral B cell differentiation, serological differences and the expression level of BCR signaling molecules between the vaccinated and recovered group. The vaccinated group exhibited reduced RBD-specific B cell differentiation and lower CD86 signal intensity on memory B cells, but enhanced BCR signaling in B cells. Regarding metabolic signaling, the vaccinated group had elevated expression levels of pS6, c-Myc, pmTOR, and pSTAT5, suggesting that the STAT5-c-Myc axis plays a role in regulating B cell metabolism. Additionally, proteome microarray analysis revealed that the serum of the vaccinated group contained higher levels of IgG antibodies against the SARS-CoV-2 N-Nter protein and IgA antibodies specific to the SARS-CoV-2 S1 protein. In summary, these findings indicate that the vaccinated group develops a more robust coronavirus-specific immune response, with enhanced BCR signaling and metabolic activity compared to the recovered group. These insights might contribute to the optimization of SARS-CoV-2 vaccine design.展开更多
With the rapid development of digital communication and the widespread use of the Internet of Things,multi-view image compression has attracted increasing attention as a fundamental technology for image data communica...With the rapid development of digital communication and the widespread use of the Internet of Things,multi-view image compression has attracted increasing attention as a fundamental technology for image data communication.Multi-view image compression aims to improve compression efficiency by leveraging correlations between images.However,the requirement of synchronization and inter-image communication at the encoder side poses significant challenges,especially for constrained devices.In this study,we introduce a novel distributed image compression model based on the attention mechanism to address the challenges associated with the availability of side information only during decoding.Our model integrates an encoder network,a quantization module,and a decoder network,to ensure both high compression performance and high-quality image reconstruction.The encoder uses a deep Convolutional Neural Network(CNN)to extract high-level features from the input image,which then pass through the quantization module for further compression before undergoing lossless entropy coding.The decoder of our model consists of three main components that allow us to fully exploit the information within and between images on the decoder side.Specifically,we first introduce a channel-spatial attention module to capture and refine information within individual image feature maps.Second,we employ a semi-coupled convolution module to extract both shared and specific information in images.Finally,a cross-attention module is employed to fuse mutual information extracted from side information.The effectiveness of our model is validated on various datasets,including KITTI Stereo and Cityscapes.The results highlight the superior compression capabilities of our method,surpassing state-of-the-art techniques.展开更多
基金supported in part by NSFC under Grant 62422407in part by RGC under Grant 26204424in part by ACCESS–AI Chip Center for Emerging Smart Systems, sponsored by the Inno HK initiative of the Innovation and Technology Commission of the Hong Kong Special Administrative Region Government
文摘Robotic computing systems play an important role in enabling intelligent robotic tasks through intelligent algo-rithms and supporting hardware.In recent years,the evolution of robotic algorithms indicates a roadmap from traditional robotics to hierarchical and end-to-end models.This algorithmic advancement poses a critical challenge in achieving balanced system-wide performance.Therefore,algorithm-hardware co-design has emerged as the primary methodology,which ana-lyzes algorithm behaviors on hardware to identify common computational properties.These properties can motivate algo-rithm optimization to reduce computational complexity and hardware innovation from architecture to circuit for high performance and high energy efficiency.We then reviewed recent works on robotic and embodied AI algorithms and computing hard-ware to demonstrate this algorithm-hardware co-design methodology.In the end,we discuss future research opportunities by answering two questions:(1)how to adapt the computing platforms to the rapid evolution of embodied AI algorithms,and(2)how to transform the potential of emerging hardware innovations into end-to-end inference improvements.
基金This research was supported in part by ACCESS-AI Chip Center for Emerging Smart Systems,sponsored by InnoHK funding,Hong Kong SAR,and HKUST-HKUST(GZ)20 for 20 Cross-campus Collaborative Research Scheme C031.
文摘Compared to the last decade when the convolution neu-ral network(CNN)dominated the research field,machine learn-ing(ML)algorithms have reached a pivotal moment called the generative artificial intelligence(AI)era.With the emer-gence of large-scale foundation models[1],such as large multi-modal model(LMM)GPT-4[2]and text-to-image generative model DALL·E[3].
文摘植物图片包含植物生境、物种组成、形态特征、物候等相关信息,是野外调查和植物记录的重要资料。无人机可按照设定程序定时、定航线拍摄植物,获取植物图片拍摄地的精准位置信息,进而实现周期化的植物拍摄和调查。本图片数据集系于2022–2023年在内蒙古呼伦贝尔湿润草原、锡林浩特典型草原、鄂尔多斯干旱草原选地,依照《草地植物多样性无人机调查技术规范》(T/CSES 123-2023)团体标准,以DJI MINI 3 PRO无人机采集而来,并以人工框选和鉴定为主、目标检测和智能识别模型处理为辅的方式进行了图像中的植物框选和鉴定。本数据集包含了19科32属40种植物的4000幅图片、植物物种名称、植物科属信息、采集时间、采集点海拔、经纬度。本数据集可以为相关草地植物的形态、分布、物候等信息检索以及智能识别模型构建提供数据支撑。
基金supported via funding from Prince Sattam bin Abdulaziz University project number(PSAU/2024/R/1445).
文摘Being a nonlinear operator,fractional derivatives can affect the enforcement of existence at any given time.As a result,the memory effect has an impact on all nonlinear processes modeled by fractional order differential equations(FODEs).The goal of this study is to increase the fractional model of the TB virus’s(FMTBV)accuracy.Stochastic solvers have never been used to solve FMTBV previously.The Bayesian regularized artificial(BRA)method and neural networks(NNs),often referred to as BRA-NNs,were used to solve the FMTBV model.Each scenario features five occurrences that each reflect a different order of derivatives,ranging from 0.8,0.85,0.9,0.95,and 1,as well as five potential rates for different parameters.Training data made up 90%of the data,testing data made up 5%,and validation data made up 5%of the data used to illustrate the FMTBV’s approximations.To verify that the BRA-NNs were correct,the generated simulations were described in the following solutions using the FOLotkaVolterra approach in MATLAB.Comprehensive Simulink results in terms of mean square error,error histogram,and regression analysis investigations further highlight the competence,dependability,and accuracy of the suggested BRA-NNs.
基金supported by grants from R&D Program of Guangzhou Laboratory(SRPG22-006)the National Natural Science Foundation of China(82371784,92374110)+2 种基金the R&D Program of Guangzhou National Laboratory(No.GZNL2023A01005)the Fourteenth Five-Year National Key Research and Development Program of China(2023YFC2307200)the State Key Laboratory for Diagnosis and Treatment of Severe Zoonotic Infectious Diseases Program(1640282).
文摘SARS-CoV-2 infection and vaccination both trigger immune responses. The former leads to naturally acquired immunity, while the latter induces active immunity through artificial means. However, the distinct immune effects of vaccination and infection, as well as their underlying mechanisms, require further clarification. In this study, we compared the peripheral B cell differentiation, serological differences and the expression level of BCR signaling molecules between the vaccinated and recovered group. The vaccinated group exhibited reduced RBD-specific B cell differentiation and lower CD86 signal intensity on memory B cells, but enhanced BCR signaling in B cells. Regarding metabolic signaling, the vaccinated group had elevated expression levels of pS6, c-Myc, pmTOR, and pSTAT5, suggesting that the STAT5-c-Myc axis plays a role in regulating B cell metabolism. Additionally, proteome microarray analysis revealed that the serum of the vaccinated group contained higher levels of IgG antibodies against the SARS-CoV-2 N-Nter protein and IgA antibodies specific to the SARS-CoV-2 S1 protein. In summary, these findings indicate that the vaccinated group develops a more robust coronavirus-specific immune response, with enhanced BCR signaling and metabolic activity compared to the recovered group. These insights might contribute to the optimization of SARS-CoV-2 vaccine design.
基金supported by the National Natural Science Foundation of China(Key Program)(No.11932013)the Tianjin Science and Technology Plan Project(No.22PTZWHZ00040)。
文摘With the rapid development of digital communication and the widespread use of the Internet of Things,multi-view image compression has attracted increasing attention as a fundamental technology for image data communication.Multi-view image compression aims to improve compression efficiency by leveraging correlations between images.However,the requirement of synchronization and inter-image communication at the encoder side poses significant challenges,especially for constrained devices.In this study,we introduce a novel distributed image compression model based on the attention mechanism to address the challenges associated with the availability of side information only during decoding.Our model integrates an encoder network,a quantization module,and a decoder network,to ensure both high compression performance and high-quality image reconstruction.The encoder uses a deep Convolutional Neural Network(CNN)to extract high-level features from the input image,which then pass through the quantization module for further compression before undergoing lossless entropy coding.The decoder of our model consists of three main components that allow us to fully exploit the information within and between images on the decoder side.Specifically,we first introduce a channel-spatial attention module to capture and refine information within individual image feature maps.Second,we employ a semi-coupled convolution module to extract both shared and specific information in images.Finally,a cross-attention module is employed to fuse mutual information extracted from side information.The effectiveness of our model is validated on various datasets,including KITTI Stereo and Cityscapes.The results highlight the superior compression capabilities of our method,surpassing state-of-the-art techniques.
