检索结果-内蒙古大学图书馆

A Dynamic Transformer Network for Vehicle Detection

IEEE Transactions on Consumer Electronics 2025年

作者： Tian, Chunwei Liu, Kai Zhang, Bob Huang, Zhixiang Lin, Chia-Wen Zhang, David Harbin Institute of Technology School of Computer Science and Technology Harbin 15001 China Macao Special Administrative Region of China PAMI Research Group University of Macau 999078 Macao Anhui University Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Key Laboratory of Electromagnetic Environmental Sensing Hefei 230601 China National Tsing Hua University Department of Electrical Engineering Institute of Communications Engineering Hsinchu Taiwan The Chinese University of Hong Kong (Shenzhen) School of Data Science Shenzhen 518172 China

Stable consumer electronic systems can assist traffic better. Good traffic consumer electronic systems require collaborative work between traffic algorithms and hardware. However, performance of popular traffic algorithms containing vehicle detection methods based on deep networks via learning data relation rather than learning differences in different lighting and occlusions is limited. In this paper, we present a dynamic Transformer network for vehicle detection (DTNet). DTNet utilizes a dynamic convolution to guide a deep network to dynamically generate weights to enhance adaptability of an obtained detector. Taking into relations of different information account, a mixed attention mechanism based channel attention and Transformer is exploited to strengthen relations of channels and pixels to extract more salient information for vehicle detection. To overcome the drawback of difference in an image account, a translation-variant convolution relies on spatial location information to refine obtained structural information for vehicle detection. Experimental results illustrate that our DTNet is competitive for vehicle detection. Code of the proposed DTNet can be obtained at https://***/hellloxiaotian/DTNet. © 2025 IEEE.

关键词： Dynamic convolution Hybrid Attention Mechanism Transformer Vehicle Detection

来源：评论

学校读者我要写书评

暂无评论

Enhanced waveform interpolative coding at 4 kbps 6

Enhanced waveform interpolative coding at 4 kbps

引用

6th IEEE Workshop on Speech Coding

作者： Gottesman, Oded Gersho, Allen Signal Compression Laboratory Department of Electrical and Computer Engineering University of California Santa BarbaraCA93106 United States

ISBN: (纸本)0780356519

This paper presents an enhanced waveform interpolative (EWI) speech coder at 4 kbps. The system incorporates novel features such as analysis-by-synthesis (AbS) vector-quantization (VQ) of the dispersion-phase, AbS optimization of the slowly evolving waveform (SEW), a special pitch search for transitions, and switched-predictive analysis-by-synthesis gain VQ. Subjective quality tests indicate that it exceeds that of MPEG-4 at 4 kbps and of G.723.1 at 5.3 kbps, and it is slightly better than that of G.723.1 at 6.3 kbps. © 1999 IEEE.

关键词： Vector quantization

来源：评论

学校读者我要写书评

暂无评论

ENHANCED ANALYSIS-BY-SYNTHESIS WAVEFORM INTERPOLATIVE CODING AT 4 KBPS 6

ENHANCED ANALYSIS-BY-SYNTHESIS WAVEFORM INTERPOLATIVE CODING...

引用

6th European Conference on Speech Communication and Technology, EUROSPEECH 1999

作者： Gottesman, Oded Gersho, Allen Signal Compression Laboratory Department of Electrical and Computer Engineering University of California Santa BarbaraCA93106 United States

This paper presents an Enhanced analysis-by-synthesis (AbS) Waveform Interpolative (EWI) speech coder at 4 kbps. The system incorporates novel features such as: AbS quantization of the slowly evolving waveform (SEW), AbS vector quantization (VQ) of the dispersion phase, a special pitch search for transitions, and switched-predictive analysis-by-synthesis gain VQ. Subjective quality tests indicate that it exceeds MPEG-4 at 4 kbps and of G.723.1 at 5.3 kbps, and it is slightly better than G.723.1 at 6.3 kbps. © 1999 6th European Conference on Speech Communication and Technology, EUROSPEECH 1999. All rights reserved.

关键词： Vector quantization

来源：评论

学校读者我要写书评

暂无评论

Multimode spectral coding of speech at 2400 bps and below

Multimode spectral coding of speech at 2400 bps and below

引用

1995 IEEE Workshop on Speech Coding for Telecommunications, SCFT 1995

作者： Das, Amitava Gersho, Allen Signal Compression Laboratory Department of Electrical and Computer Engineering University of California Santa BarbaraCA93106 United States

We present a low bit rate multimode speech coding algorithm which applies a suitable spectral model and a custom quantization scheme to each frame according to the selected mode. For unvoiced speech, the spectrum is characterized by a fixed-dimension spectral shape vector (SSV), while the spectra of 13 different voiced modes are specified by the pitch and a variable dimension SSV. A variable dimension vector quantization technique offers efficient spectral quantization and improves overall speech quality. A variable rate (1.7 kb/s avg. rate;2.4 kb/s max. rate) implementation of the multimode algorithm yields speech quality comparable to the 4.8 kb/s FS 1016 CELP coder and the 4.15 kb/s Inmarsat-MIMBE coder. © 1995 Proceedings - 1995 IEEE Workshop on Speech Coding for Telecommunications: Speech Coding for Interoperable Global Communications, SCFT 1995. All rights reserved.

关键词： Speech coding

来源：评论

学校读者我要写书评

暂无评论

compression of multifractal object boundaries: A comparison of two approaches

Compression of multifractal object boundaries: A comparison ...

引用

Canadian Conference on electrical and computer engineering 2005

作者： Siddiqui, S. El-Boustani, A. Kinsner, W. Signal and Data Compression Laboratory Department of Electrical and Computer Engineering University of Manitoba Winnipeg Man. R3T 5V6 Canada

This paper presents a comparative study of two approaches that address the problem of compression of arbitrary object boundaries resulting from image segmentation. The first approach uses multiresolution based techniques, in which the boundary is represented compactly by two sets of descriptors and control points which are obtained by applying wavelet and fractal analysis. The midpoint displacement algorithm is then applied to the two sets of control points in order to reconstruct the boundaries with the required fractal dimension. The second approach is based on the iterated function system (IFS). In this approach, the object boundary is regarded as an attractor of an IFS having the same Rényi fractal dimension spectrum as the original boundary. The experimental results show that the multifractal object boundaries can be modelled compactly using both approaches. The compression ratio achieved for the wavelet-based approach was 285:1, while for the IFS-based approach it was 52:1. These results can be compared with state of the art image compression techniques such as JPEG2000. © 2005 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Surrogate ECG signals for the null-hypothesis analysis of heart monitoring

Surrogate ECG signals for the null-hypothesis analysis of he...

引用

Canadian Conference on electrical and computer engineering 2005

作者： Potter, M. Kinsner, W. Department of Electrical and Computer Engineering Signal and Data Compression Laboratory University of Manitoba Winnipeg Man. R3T 5V6 Canada

This paper introduces a class of synthetic electrocardiograms (ECG) for the purpose of testing the sensitivity and robustness of heart-monitoring techniques in controlled and reproducible simulations. These ECG surrogates are constructed by adapting an existing standard ECG synthesizer to act as a modulator on a tachogram input, enabling null-hypothesis analysis on heart-rate variability. This establishes an empirical link between the tachogram literature and the design of heart monitoring techniques. The new methodology is demonstrated experimentally and its performance is discussed. © 2005 IEEE.

关键词： Electrocardiography

来源：评论

学校读者我要写书评

暂无评论

Rotation sensitivity of independent component analysis to outliers

Rotation sensitivity of independent component analysis to ou...

引用

Canadian Conference on electrical and computer engineering 2005

作者： Gadhok, N. Kinsner, W. Department of Electrical and Computer Engineering Signal and Data Compression Laboratory University of Manitoba Winnipeg Man. R3T 5V6 Canada

This paper evaluates the outlier sensitivity of five independent component analysis (ICA) algorithms (FastICA, Extended-Infomax, JADE, Radical-ICA, and β-divergence) using (i) the Amari separation performance index, (ii) the optimum angle of rotation error, and (iii) the contrast function difference, in an outlier-contaminated mixture simulation. The Amari separation performance index has revealed a strong sensitivity of JADE and FastICA (using 3rd- and 4th-order nonlinearities) to outliers. However, the two contrast measures demonstrated conclusively that β-divergence is the least outlier-sensitive algorithm, followed by Radical-ICA, FastICA (exponential and hyperbolic-tangent nonlinearities), Extended-Infomax, JADE, and FastICA (3rd- and 4th-order nonlinearities) in an outlier-contaminated mixture of two uniformly distributed signals. The novelty of this paper is the development of an unbiased optimization-landscape environment for assessing outlier sensitivity, as well as the optimum angle of rotation error and the contrast function difference as promising new measures for assessing the outlier sensitivity of ICA algorithms. © 2005 IEEE.

关键词： Independent component analysis

来源：评论

学校读者我要写书评

暂无评论

Local and global analysis of multifractal singularity spectrum through wavelets

Local and global analysis of multifractal singularity spectr...

引用

Canadian Conference on electrical and computer engineering 2005

作者： Faghfouri, A. Kinsner, W. Dept. of Electrical and Computer Engineering Signal and Data Compression Laboratory University of Manitoba Winnipeg Man. R3T 5V6 Canada

This paper presents the implementation aspects pertinent to the computation of the multifractal singularity spectrum through wavelets, and the methods of overcoming them. Multifractals are mixtures of monofractals, and monofractals are self-affine objects that hold power law relationships over several scales. Such multifractals can be detected and measured through a singularity spectrum. Many natural and artificial phenomena such as turbulence, diffusion limited aggregates, and electrical discharges exhibit multifractality. Since these phenomena are highly nonlinear and nonstationary, regular analyses such as Fourier decomposition cannot characterize them effectively. In order to characterize, compare and quantify multifractal objects, appropriate measures such as the Rényi fractal dimension spectrum (RS) and Mandelbrot singularity spectrum (MS) are required. There are two major methods for calculating a singularity spectrum;one is through Legendre transform of the RS, and the other is through the wavelet transform modulus maxima (WTMM) method. This paper provides solutions to the difficulties that arise in the computation of the MS through WTMM, for one-dimensional signals and compares them to the existing multifractal literature and software implementations. Appropriate mother wavelets, continuous wavelet implementation, and thresholding of the wavelet coefficients are also discussed. © 2005 IEEE.

关键词： signal processing

来源：评论

学校读者我要写书评

暂无评论

Singularity processing of nonstationary signals

Singularity processing of nonstationary signals

引用

Canadian Conference on electrical and computer engineering (CCECE)

作者： A. Langi W. Kinsner Department of Electrical and Computer Engineering Signal and Data Compression Laboratory University of Manitoba MAN Canada

This paper presents a new approach in processing nonstationary signals-such as speech signals and images-through singularity characterization. In this approach, we associate a singular measure /spl mu//sub f(t/) (r) with a transient at time t of a signal f(t) (where a real number r>0 is a time perturbation around t) and use the singularity behaviour of the measure for the characterization of the signal nonstationarity. The approach is capable of characterizing isolated transients through Holder exponents (or singularity strength), as well as mixture transients (e.g. singularity everywhere) through the concept of fractality and multifractality. The paper discusses the concept and the practicality of applying this approach to signals. The paper also shows that this approach can provide a unifying framework for previously published work on applying nonlinear, chaotic, fractal, and multifractal analysis to signals. We show that the main conceptual issue in applying fractality and multifractality to signals using this framework is the proper selection of signal measures.

关键词： signal processing Fractals signal analysis Speech analysis Chaos Image edge detection Image coding Speech recognition Image analysis Image texture analysis

来源：评论

学校读者我要写书评

暂无评论

Secure and resilient data printed on paper

Secure and resilient data printed on paper

引用

Canadian Conference on electrical and computer engineering (CCECE)

作者： A. Denis W. Kinsner Department of Electrical and Computer Engineering Signal and Data Compression Laboratory University of Manitoba Winnipeg MAN Canada

Transmission and printing of sensitive information requires both data security, and protection against random and burst errors. This paper describes a technique that achieves two objectives: secure and reliable transmission of information; and integrity of the original information printed on paper. Information printed on paper is prone to burst and random errors. Resilience of the information to such errors is obtained by introducing redundancy into the original data using forward error correcting codes. We have selected a concatenation of a Reed-Solomon (RS) code, interleaved with a self-orthogonal majority decodable convolutional code. To measure the performance of the concatenated code, the bit error rate against signal to noise ratio of the code is compared with codes of equivalent rate (e.g., RS code alone) and with the unencoded data. Security of the printed data can be achieved through encryption. We have selected a probabilistic encryption scheme of messages to achieve an increased security against reverse engineering of the printed pattern. A new document signature extraction scheme based on fractal signal processing is described. Such a signature is then included with the document in the security pattern.

关键词： data security Error correction codes Cryptography Printing Protection Resilience Redundancy Reed-Solomon codes Decoding Convolutional codes

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：