检索结果-内蒙古大学图书馆

Projected residual vector quantization for ANN Search

IEEE MULTIMEDIA 2014年第3期21卷 41-51页

作者： Wei, Benchang Guan, Tao Yu, Junqing Huazhong Univ Sci & Technol Sch Comp Sci & Technol Wuhan Peoples R China

In this research, we propose Projected residual vector quantization (PRVQ) to deal with the problem of large-scale approximate nearest neighbor (ANN) search in a high-dimensional space. A lot of quantization-based ANN search algorithms have been proposed in the past few years. However, most of the existing methods discard the projection errors generated in the dimension reduction process, which inevitably decreases the search accuracy. In view of that, the authors propose a method of projected residual vector quantization for ANN search that considers the projection errors in the quantization process. They also design three simple and effective optimization strategies to improve the performance of the PRVQ algorithm. The authors have integrated the proposed PRVQ algorithm into a mobile landmark recognition system to prove its effectiveness.

关键词： Accuracy Approximation algorithms Artificial neural networks Principal component analysis quantization (signal) Search problems vectors approximate nearest neighbor search asymmetric distance high dimensional multimedia residual vector quantization Approximation algorithms tabu search recognition system Quantized Image MULTI-MEDIA SYSTEM vector quantization Nearest neighbor query

来源：评论

学校读者我要写书评

暂无评论

A fast PNN design algorithm for entropy-constrained residual vector quantization

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 1998年第7期7卷 1045-1050页

作者： Kossentini, F Smith, MJT Univ British Columbia Dept Elect & Comp Engn Vancouver BC V6T 1Z4 Canada Georgia Inst Technol Sch Elect & Comp Engn Atlanta GA 30332 USA

A clustering algorithm based on the pairwise nearest-neighbor algorithm developed by Equitz, is introduced for the design of entropy-constrained residual vector quantizers. The algorithm designs residual vector quantization codebooks by merging the pair of stage clusters that minimizes the increase in overall distortion subject to a given decrease in entropy. Image coding experiments show that the clustering design algorithm typically results in more than a 200:1 reduction in design time relative to the standard iterative entropy-constrained residual vector quantization algorithm while introducing only small additional distortion. Multipath searching over the sequence of merges is also investigated and shown experimentally to slightly improve rate-distortion performance, The proposed algorithm can be used alone or can be followed by the iterative algorithm to improve the reproduction quality at the same bit rate.

关键词： image coding entropy-constrained residual vector quantization multipath searching pairwise nearest neighbor algorithm residual vector quantization

来源：评论

学校读者我要写书评

暂无评论

Approximate Nearest Neighbor Search by residual vector quantization

引用

SENSORS 2010年第12期10卷 11259-11273页

作者： Chen, Yongjian Guan, Tao Wang, Cheng Huazhong Univ Sci & Technol Sch Comp Sci & Technol Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Digital Engn & Simulat Res Ctr Wuhan 430074 Peoples R China

A recently proposed product quantization method is efficient for large scale approximate nearest neighbor search, however, its performance on unstructured vectors is limited. This paper introduces residual vector quantization based approaches that are appropriate for unstructured vectors. Database vectors are quantized by residual vector quantizer. The reproductions are represented by short codes composed of their quantization indices. Euclidean distance between query vector and database vector is approximated by asymmetric distance, i.e., the distance between the query vector and the reproduction of the database vector. An efficient exhaustive search approach is proposed by fast computing the asymmetric distance. A straight forward non-exhaustive search approach is proposed for large scale search. Our approaches are compared to two state-of-the-art methods, spectral hashing and product quantization, on both structured and unstructured datasets. Results show that our approaches obtain the best results in terms of the trade-off between search quality and memory usage.

关键词： approximate nearest neighbor search high-dimensional indexing residual vector quantization

来源：评论

学校读者我要写书评

暂无评论

Parametric and nonparametric residual vector quantization optimizations for ANN search

引用

NEUROCOMPUTING 2016年 217卷 92-102页

作者： Guo, Dan Li, Chuanqing Wu, Lv Hefei Univ Technol Sch Comp & Informat Hefei Peoples R China Wuhan Univ Technol Sch Informat Engn Wuhan Peoples R China

For approximate nearest neighbor (ANN) search in many vision-based applications, vector quantization (VQ) is an efficient compact encoding technology. A representative approach of VQ is product quantization (PQ) which quantizes subspaces separately by Cartesian product and achieves high accuracy. But its space decomposition still leads to quantization distortion. This paper presents two optimized solutions based on residual vector quantization (RVQ). Different from PQ RVQ simulates restoring quantization error by multi-stage quantizers instead of decomposing it. To further optimize codebook and space decomposition, we try to get a better discriminated space projection. Then an orthonormal matrix R is generated. The RVQ's nonparametric solution alternately optimizes R and stage-codebooks by Singular Value Decomposition (SVD) in multiple iterations. The RVQs parametric solution assumes that data are subject to Gaussian distribution and uses Eigenvalue Allocation to get each, stage-matrix (R-1} (1 <= l <= L) at once, where L is the stage number of RVQ. Compared to various optimized PQ-based methods, our methods have good superiority on restoring quantization error. (C) 2016 Elsevier B.V. All rights reserved.

关键词： residual vector quantization vector quantization optimization Parametric optimization Nonparametric optimization Stage codebook

来源：评论

学校读者我要写书评

暂无评论

ERVQ: Leverage residual vector quantization for Speech Emotion Recognition 14

ERVQ: Leverage Residual Vector Quantization for Speech Emoti...

引用

14th International Symposium on Chinese Spoken Language Processing

作者： Xie, Jingran Xiang, Yang Wang, Hui Wu, Xixin Wu, Zhiyong Meng, Helen Tsinghua Univ Shenzhen Int Grad Sch Shenzhen Peoples R China Pengcheng Lab Shenzhen Peoples R China Chinese Univ Hong Kong Hong Kong Peoples R China

ISBN: (纸本)9798331516833;9798331516826

Although speech pre-trained models (PTM) have shown remarkable performance in speech emotion recognition (SER), they are constructed for general tasks and exhibit limitations in capturing emotional related features. Recently, some popular text-to-speech models utilize residual vector quantization (RVQ) to effectively embed multi-scale speech detail information, resulting in significant speech reconstruction quality. Inspired by this success, we explore RVQ's potential for emotional representation learning. To enhance ERVQ's ability to capture emotional features, we adopt modification and employ content alignment on it. In this paper, we present a novel perspective on SER by introducing enhanced RVQ for emotion recognition, called Emotion RVQ (ERVQ). Based on experimental results, our approach achieves State-of-the-Art (SOTA) in both in-domain and out-of-domain SER scenarios, proving its outperformed emotional representation embedding demonstrating compared to other speech PTMs.

关键词： speech emotion recognition residual vector quantization

来源：评论

学校读者我要写书评

暂无评论

Mode Based K-Means Algorithm with residual vector quantization for Compressing Images

Mode Based K-Means Algorithm with Residual Vector Quantizati...

引用

1st International Conference on Logic, Information, Control and Computation

作者： Somasundaram, K. Rani, M. Mary Shanthi Deemed Univ Gandhigram Rural Inst Dept Comp Sci & Applicat Image Proc Lab Gandhigram Tamil Nadu India

ISBN: (纸本)9783642192623

Image compression plays a vital role in many online applications like Video Conferencing, High Definition Television, Satellite Communication and other applications that demand fast and massive transmission of images. In this paper, we propose a Mode based K-means method that combines K-Means and residual vector quantization(RVQ) for compressing images. Three processes are involved in this approach;Partitioning and Clustering, Pruning and Construction of Master codebook and residual vector quantization. Extensive experiments show that this method obtains a fast solution with better compression rate and comparable PSNR than conventional K-Means algorithm.

关键词： K-Means Mode Median residual vector quantization Break Even Point Master Codebook residual Codebook

来源：评论

学校读者我要写书评

暂无评论

TARGET TRACKING USING residual vector quantization

TARGET TRACKING USING RESIDUAL VECTOR QUANTIZATION

引用

International Conference on Digital Image Computing Techniques and Applications (DICTA)

作者： Aslam, Salman Barnes, Christopher Bobick, Aaron Natl Univ Sci & Technol Islamabad Pakistan Georgia Inst Technol Atlanta GA USA

ISBN: (纸本)9781467321808;9781467321792

In this work, our goal is to track visual targets using residual vector quantization (RVQ). We compare our results with principal components analysis (PCA) and tree structured vector quantization (TSVQ) based tracking. This work is significant since PCA is commonly used in the Pattern Recognition, Machine Learning and Computer Vision communities. On the other hand, TSVQ is commonly used in the Signal Processing and data compression communities. RVQ with more than two stages has not received much attention due to the difficulty in producing stable designs. In this work, we bring together these different approaches into an integrated tracking framework and show that RVQ tracking performs best according to multiple criteria over a variety of publicly available datasets. Moreover, an advantage of our approach is a learning-based tracker that builds the target model while it tracks, thus avoiding the costly step of building target models prior to tracking.

关键词： residual vector quantization tracking RVQ PCA TSVQ learning generalization

来源：评论

学校读者我要写书评

暂无评论

TSP-TTS: Text-based Style Predictor with residual vector quantization for Expressive Text-to-Speech 25

TSP-TTS: Text-based Style Predictor with Residual Vector Qua...

引用

25th Interspeech Conference

作者： Seong, Donghyun Lee, Hoyoung Chang, Joon-Hyuk Hanyang Univ Dept Elect Engn Seoul South Korea Hanyang Univ Dept Artificial Intelligence Applicat Seoul South Korea

Expressive text-to-speech (TTS) aims to synthesize better human-like speech by incorporating diverse speech styles or emotions. While most expressive TTS models rely on reference speech to condition the style of the generated speech, they often fail to generate speech of regular quality. To ensure consistent speech quality, we propose an expressive TTS conditioned on style representation extracted from the text itself. To implement this text-based style predictor, we design a style module incorporating residual vector quantization. Furthermore, the style representation is enhanced through style-to-text alignment and a mel decoder with style hierarchical layer normalization (SHLN). Our experimental findings demonstrate that our proposed model accurately estimates style representation, enabling the generation of high-quality speech without the need for reference speech.

关键词： Text-to-speech expressive speech synthesis residual vector quantization

来源：评论

学校读者我要写书评

暂无评论

Spatio-Temporal Focal Modulation Network for Massive MIMO CSI Feedback

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2025年第3期74卷 5259-5263页

作者： Shao, Hua Zhang, Xiao Zhang, Haijun Xu, Kexin Univ Sci & Technol Beijing Inst Artificial Intelligence Sch Intelligence Sci & Technol Beijing 100083 Peoples R China Univ Sci & Technol Beijing Sch Comp & Commun Engn Beijing 100083 Peoples R China

Channel state information (CSI) compression and recovery are vital for the communication system to maximize the throughput. In this paper, a CSI feedback method based on spatio-temporal focal modulation (STFM) is proposed. By using a focal modulation mechanism that incorporates the interaction between the spatial and temporal features, the proposed STFM jointly exploits the cross-domain correlation. With multiple successive CSI snapshots, STFM can achieve excellent compression and reconstruction performance. Besides, a residual vector quantization (RVQ) with a shared and learnable codebook is introduced. The RVQ includes multiple stages to quantize the output of the encoder recursively and the corresponding quantization error can be reduced. Simulation results show that the STFM with RVQ outperforms the existing CSI feedback compression and quantization methods in different configurations.

关键词： Tensors Decoding Convolution Modulation Massive MIMO Feature extraction Downlink Correlation vectors vector quantization Spatio-temporal focal modulation residual vector quantization massive MIMO CSI feedback

来源：评论

学校读者我要写书评

暂无评论

Late-season rural land-cover estimation with polarimetric-SAR intensity pixel blocks and σ-tree-structured near-neighbor classifiers

引用

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING 2006年第9期44卷 2384-2392页

作者： Barnes, Christopher F. Burki, Jehanzeb Georgia Inst Technol Sch Elect & Comp Engn Savannah GA 31407 USA

Synthetic aperture radar (SAR) image classification for late-season rural land-cover estimation is investigated. A novel tree-structured nearest neighbor-like classifier is applied to polarimetric SAR intensity image pixel blocks. The novel tree structure, called a sigma-tree, is generated by an ordered summation of unweighted template refinements. Computation and memory costs of a sigma-tree classifier grow linearly. The reduced costs of sigma-tree classifiers are obtained with the tradeoff of a guarantee of nearest neighbor mappings. Causal-anticausal refinement-template design methods, combined with causal multiple-stage search engine structures, are shown to yield sequential search decisions that are acceptably near-neighbor mappings. The performance of a sigma-tree classifier is demonstrated for rural land-cover estimation with detected polarimetric C-band AirSAR pixel data. Experiments are conducted on various polarization/pixel block size combinations to evaluate the relative utility of spatial-only, polarimetric-only, and combined spatial/polarimetric classifier inputs.

关键词： additive successive refinements direct sum residual vector quantization successive approximation synthetic aperture radar (SAR) land-use classification sigma-trees

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：