With the increasing application of surveillance cameras,vehicle re-identication(Re-ID)has attracted more attention in the eld of public security.Vehicle Re-ID meets challenge attributable to the large intra-class diff...With the increasing application of surveillance cameras,vehicle re-identication(Re-ID)has attracted more attention in the eld of public security.Vehicle Re-ID meets challenge attributable to the large intra-class differences caused by different views of vehicles in the traveling process and obvious inter-class similarities caused by similar appearances.Plentiful existing methods focus on local attributes by marking local locations.However,these methods require additional annotations,resulting in complex algorithms and insufferable computation time.To cope with these challenges,this paper proposes a vehicle Re-ID model based on optimized DenseNet121 with joint loss.This model applies the SE block to automatically obtain the importance of each channel feature and assign the corresponding weight to it,then features are transferred to the deep layer by adjusting the corresponding weights,which reduces the transmission of redundant information in the process of feature reuse in DenseNet121.At the same time,the proposed model leverages the complementary expression advantages of middle features of the CNN to enhance the feature expression ability.Additionally,a joint loss with focal loss and triplet loss is proposed in vehicle Re-ID to enhance the model’s ability to discriminate difcult-to-separate samples by enlarging the weight of the difcult-to-separate samples during the training process.Experimental results on the VeRi-776 dataset show that mAP and Rank-1 reach 75.5%and 94.8%,respectively.Besides,Rank-1 on small,medium and large sub-datasets of Vehicle ID dataset reach 81.3%,78.9%,and 76.5%,respectively,which surpasses most existing vehicle Re-ID methods.展开更多
To identify the unsteady°uid dynamic model for a solid moving in the°uid,a new identication method that operates in the continuous time domain is proposed.It is illustrated that this new identication method ...To identify the unsteady°uid dynamic model for a solid moving in the°uid,a new identication method that operates in the continuous time domain is proposed.It is illustrated that this new identication method is more accurate and appropriate for identifying unsteady°uid dynamic model in incompressible°ow in comparison with the classical frequency domain or discrete domain method.This new method is applied to identify the unsteady°uid dynamic models of a NACA0012 airfoil and the e®ect of the°uid viscosity on the model parameter is analyzed.The result shows that the added mass in viscous°ow is dependent on the input signal frequency and is exactly the added mass in inviscid°ow when the frequency approaches innity.Based on this discovery,the long controversy on the relationship of added mass between viscous and inviscid°ow is solved.Then,the enlightenment of this discovery on the identication of both linear and nonlinear unsteady aerodynamic model in incompressible°ow is illustrated.At last,the advantage of the new acquired unsteady°uid dynamic model is discussed in comparison with the classical quasi-steady aerodynamic model.展开更多
Signals from multi-sensor systems are often mixtures of （statistically） independent sources by unknown mixing method. Blind source separation（BSS） and independent component analysis（ICA） are the methods to ident...Signals from multi-sensor systems are often mixtures of （statistically） independent sources by unknown mixing method. Blind source separation（BSS） and independent component analysis（ICA） are the methods to identify/recover the channels and the sources. BSS/ICA of nonlinear mixing models are difficult problems. For instance, the post-nonlinear model has been studied by several authors. It is noticed that in most cases, the proposed models are always with an invertible mixing. According to this fact there is an interesting question, how about the situation of the non-invertible non-linear mixing in BSS or ICA？ A new simple non-linear mixing model is proposed with a kind of non-invertible mixing, the folding mixing, and method to identify its channel, blindly.展开更多
The Bohai Bay Basin is a Meso-Cenozoic rifted basin where the Paleozoic buried hills with great hydrocarbon potentials are well developed. The reservoir space types are complex and diverse due to tectonic activities, ...The Bohai Bay Basin is a Meso-Cenozoic rifted basin where the Paleozoic buried hills with great hydrocarbon potentials are well developed. The reservoir space types are complex and diverse due to tectonic activities, making fracture distribution highly heterogeneous. Reservoir identification and mapping is challenging due to their large burial depth and poor resolution of seismic data. An integration of well-logging, seismic data interpretation and core observation is applied to identify three structural unit types in the study area, that is, fault breccia zone, fault cataclastic zone, and fault massive rock zone. A comprehensive well-logging identification template and a comprehensive discriminant function M for the reservoir are established based on the well-logging response characteristics. A M value greater than 0.12 indicates a fault breccia zone, that between 0.04 and 0.12 marks a fault cataclastic zone, and that in the range from 0.02 to 0.04 represents a fault massive rock zone. A seismic prediction method with multi-parameter fusion is proposed in the study. The large-scale fractures are mapped by coherence-clutter parameters, while small fractures are predicted via waveform indication inversion. The spatial distribution of “fault-fracture reservoirs” is precisely mapped by frequency fusion technology. It is found that the fault breccia zones usually occur close to the fault planes, while the fault cataclastic zones are slightly away from the fault planes. The hydrocarbon abundance of the breccia zones is greater than that of the fault cataclastic and fault massive rock zones.展开更多
The inverse problem of reconstructing the time-dependent thermal conductivity and free boundary coefcients along with the temperature in a two-dimensional parabolic equation with initial and boundary conditions and ad...The inverse problem of reconstructing the time-dependent thermal conductivity and free boundary coefcients along with the temperature in a two-dimensional parabolic equation with initial and boundary conditions and additional measurements is,for the rst time,numerically investigated.This inverse problem appears extensively in the modelling of various phenomena in engineering and physics.For instance,steel annealing,vacuum-arc welding,fusion welding,continuous casting,metallurgy,aircraft,oil and gas production during drilling and operation of wells.From literature we already know that this inverse problem has a unique solution.However,the problem is still ill-posed by being unstable to noise in the input data.For the numerical realization,we apply the alternating direction explicit method along with the Tikhonov regularization to nd a stable and accurate numerical solution of nite differences.The root mean square error(rmse)values for various noise levels p for both smooth and non-smooth continuous time-dependent coef-cients Examples are compared.The resulting nonlinear minimization problem is solved numerically using the MATLAB subroutine lsqnonlin.Both exact and numerically simulated noisy input data are inverted.Numerical results presented for two examples show the efciency of the computational method and the accuracy and stability of the numerical solution even in the presence of noise in the input data.展开更多
Steganography has been used to prevent unauthorized access to private information during transmission.It is the scheme of securing sensitive information by concealing it within carriers such as digital images,videos,a...Steganography has been used to prevent unauthorized access to private information during transmission.It is the scheme of securing sensitive information by concealing it within carriers such as digital images,videos,audio,or text.Current steganography methods are working by assigning a cover le then embed the payload within it by making some modications,creating the stego-le.However,the left traces that are caused by these modications will make steganalysis algorithms easily detect the hidden payload.Aiming to solve this issue,a novel,highly robust steganography method based on hacking anti-shoplifting systems has proposed to hide data within clothes.The anti-Shoplifting system is an anti-theft security system that protects goods and products,leaving the store in an illegal way(i.e.,without paying for them).The proposed method works by modifying the default anti-shoplifting system by changing its built-in soft RFID(radio-frequency identication)tags sewn in clothes into NFC(Near Field Communication)tags.These NFC tags are smart tags that can communicate with NFC-Enabled smart-phones using NDEF(NFC Data Exchange Format).NDEF is one of the advancements added to RFID technology by NFC,which allows the data exchange.Every NDEF message has one/more NDEF records that contain record type,a unique ID,a length,and a payload of data that contains the secret message content that can be any type of data that ts in a byte stream.Based on NDEF and NFC-enabled smart-phones,the proposed method will take the secret message from the sender,make use of his NFC-enabled smart-phone to communicate with the NFC tag,then hide the secret message within the NDEF’s payload of the NFC tag stuck in clothes.Finally,to evaluate the proposed method,it has been compared with default(digital)steganography weak points.Such as time,lockable,robustness,attacks,capacity,and a few more points.The results and comparisons showed that the proposed method is more efcient than default(digital)steganography and has many advantages.展开更多
基金supported,in part,by the National Nature Science Foundation of China under Grant Numbers 61502240,61502096,61304205,61773219in part,by the Natural Science Foundation of Jiangsu Province under Grant Numbers BK20201136,BK20191401in part,by the Priority Academic Program Development of Jiangsu Higher Education Institutions(PAPD)fund.
文摘With the increasing application of surveillance cameras,vehicle re-identication(Re-ID)has attracted more attention in the eld of public security.Vehicle Re-ID meets challenge attributable to the large intra-class differences caused by different views of vehicles in the traveling process and obvious inter-class similarities caused by similar appearances.Plentiful existing methods focus on local attributes by marking local locations.However,these methods require additional annotations,resulting in complex algorithms and insufferable computation time.To cope with these challenges,this paper proposes a vehicle Re-ID model based on optimized DenseNet121 with joint loss.This model applies the SE block to automatically obtain the importance of each channel feature and assign the corresponding weight to it,then features are transferred to the deep layer by adjusting the corresponding weights,which reduces the transmission of redundant information in the process of feature reuse in DenseNet121.At the same time,the proposed model leverages the complementary expression advantages of middle features of the CNN to enhance the feature expression ability.Additionally,a joint loss with focal loss and triplet loss is proposed in vehicle Re-ID to enhance the model’s ability to discriminate difcult-to-separate samples by enlarging the weight of the difcult-to-separate samples during the training process.Experimental results on the VeRi-776 dataset show that mAP and Rank-1 reach 75.5%and 94.8%,respectively.Besides,Rank-1 on small,medium and large sub-datasets of Vehicle ID dataset reach 81.3%,78.9%,and 76.5%,respectively,which surpasses most existing vehicle Re-ID methods.
文摘To identify the unsteady°uid dynamic model for a solid moving in the°uid,a new identication method that operates in the continuous time domain is proposed.It is illustrated that this new identication method is more accurate and appropriate for identifying unsteady°uid dynamic model in incompressible°ow in comparison with the classical frequency domain or discrete domain method.This new method is applied to identify the unsteady°uid dynamic models of a NACA0012 airfoil and the e®ect of the°uid viscosity on the model parameter is analyzed.The result shows that the added mass in viscous°ow is dependent on the input signal frequency and is exactly the added mass in inviscid°ow when the frequency approaches innity.Based on this discovery,the long controversy on the relationship of added mass between viscous and inviscid°ow is solved.Then,the enlightenment of this discovery on the identication of both linear and nonlinear unsteady aerodynamic model in incompressible°ow is illustrated.At last,the advantage of the new acquired unsteady°uid dynamic model is discussed in comparison with the classical quasi-steady aerodynamic model.
基金This project was supported by the Talent Foundation of Anhui Province(2004Z025)
文摘Signals from multi-sensor systems are often mixtures of （statistically） independent sources by unknown mixing method. Blind source separation（BSS） and independent component analysis（ICA） are the methods to identify/recover the channels and the sources. BSS/ICA of nonlinear mixing models are difficult problems. For instance, the post-nonlinear model has been studied by several authors. It is noticed that in most cases, the proposed models are always with an invertible mixing. According to this fact there is an interesting question, how about the situation of the non-invertible non-linear mixing in BSS or ICA？ A new simple non-linear mixing model is proposed with a kind of non-invertible mixing, the folding mixing, and method to identify its channel, blindly.
文摘The Bohai Bay Basin is a Meso-Cenozoic rifted basin where the Paleozoic buried hills with great hydrocarbon potentials are well developed. The reservoir space types are complex and diverse due to tectonic activities, making fracture distribution highly heterogeneous. Reservoir identification and mapping is challenging due to their large burial depth and poor resolution of seismic data. An integration of well-logging, seismic data interpretation and core observation is applied to identify three structural unit types in the study area, that is, fault breccia zone, fault cataclastic zone, and fault massive rock zone. A comprehensive well-logging identification template and a comprehensive discriminant function M for the reservoir are established based on the well-logging response characteristics. A M value greater than 0.12 indicates a fault breccia zone, that between 0.04 and 0.12 marks a fault cataclastic zone, and that in the range from 0.02 to 0.04 represents a fault massive rock zone. A seismic prediction method with multi-parameter fusion is proposed in the study. The large-scale fractures are mapped by coherence-clutter parameters, while small fractures are predicted via waveform indication inversion. The spatial distribution of “fault-fracture reservoirs” is precisely mapped by frequency fusion technology. It is found that the fault breccia zones usually occur close to the fault planes, while the fault cataclastic zones are slightly away from the fault planes. The hydrocarbon abundance of the breccia zones is greater than that of the fault cataclastic and fault massive rock zones.
文摘The inverse problem of reconstructing the time-dependent thermal conductivity and free boundary coefcients along with the temperature in a two-dimensional parabolic equation with initial and boundary conditions and additional measurements is,for the rst time,numerically investigated.This inverse problem appears extensively in the modelling of various phenomena in engineering and physics.For instance,steel annealing,vacuum-arc welding,fusion welding,continuous casting,metallurgy,aircraft,oil and gas production during drilling and operation of wells.From literature we already know that this inverse problem has a unique solution.However,the problem is still ill-posed by being unstable to noise in the input data.For the numerical realization,we apply the alternating direction explicit method along with the Tikhonov regularization to nd a stable and accurate numerical solution of nite differences.The root mean square error(rmse)values for various noise levels p for both smooth and non-smooth continuous time-dependent coef-cients Examples are compared.The resulting nonlinear minimization problem is solved numerically using the MATLAB subroutine lsqnonlin.Both exact and numerically simulated noisy input data are inverted.Numerical results presented for two examples show the efciency of the computational method and the accuracy and stability of the numerical solution even in the presence of noise in the input data.
文摘Steganography has been used to prevent unauthorized access to private information during transmission.It is the scheme of securing sensitive information by concealing it within carriers such as digital images,videos,audio,or text.Current steganography methods are working by assigning a cover le then embed the payload within it by making some modications,creating the stego-le.However,the left traces that are caused by these modications will make steganalysis algorithms easily detect the hidden payload.Aiming to solve this issue,a novel,highly robust steganography method based on hacking anti-shoplifting systems has proposed to hide data within clothes.The anti-Shoplifting system is an anti-theft security system that protects goods and products,leaving the store in an illegal way(i.e.,without paying for them).The proposed method works by modifying the default anti-shoplifting system by changing its built-in soft RFID(radio-frequency identication)tags sewn in clothes into NFC(Near Field Communication)tags.These NFC tags are smart tags that can communicate with NFC-Enabled smart-phones using NDEF(NFC Data Exchange Format).NDEF is one of the advancements added to RFID technology by NFC,which allows the data exchange.Every NDEF message has one/more NDEF records that contain record type,a unique ID,a length,and a payload of data that contains the secret message content that can be any type of data that ts in a byte stream.Based on NDEF and NFC-enabled smart-phones,the proposed method will take the secret message from the sender,make use of his NFC-enabled smart-phone to communicate with the NFC tag,then hide the secret message within the NDEF’s payload of the NFC tag stuck in clothes.Finally,to evaluate the proposed method,it has been compared with default(digital)steganography weak points.Such as time,lockable,robustness,attacks,capacity,and a few more points.The results and comparisons showed that the proposed method is more efcient than default(digital)steganography and has many advantages.
