2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS)最新文献_第8页

Thermometer Code of Log Mel-Frequency Spectral Coefficient for BNN-based Keyword Spotting System Log mel温度计代码-基于bnn的关键词识别系统的频谱系数

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090286

Yuzhong Jiao, Yiu Kei Li, C. Chan, Yun Li, Zhilin Ai

{"title":"Thermometer Code of Log Mel-Frequency Spectral Coefficient for BNN-based Keyword Spotting System","authors":"Yuzhong Jiao, Yiu Kei Li, C. Chan, Yun Li, Zhilin Ai","doi":"10.1109/APCCAS55924.2022.10090286","DOIUrl":"https://doi.org/10.1109/APCCAS55924.2022.10090286","url":null,"abstract":"For keyword spotting (KWS) systems that usually work in mobile devices, a low-complexity design is essential for long stand-by time. Audio feature extraction and classifier modeling are the two main components of KWS systems. Log Mel-Frequency Spectral Coefficient (MFSC) is common for audio feature extraction due to its low complexity and good performance. Binary neural network (BNN) classifier, which owns binary weights and activations and performs convolution with XNOR, is applicable to low-complexity KWS applications. However, audio features are usually quantized with multiple-bit binary code to maintain high classification accuracy, which requires addition (ADD) operations in the first convolutional layer of the BNN model. Therefore, both XNOR and ADD units are needed in the BNN accelerator. To further reduce the complexity of KWS systems, we propose a new feature extraction method: Thermometer Codes of MFSC (MFSC-TC). Without LOG and DELTA operations, it is simpler than other MFSC-based methods. More importantly, convolution of all layers can be done by XNOR units due to the feature of thermometer code. The experiments with the Google Speech Commands dataset validate that the MFSC-TC-based BNN models outperform the models with more layers using other feature extraction methods.","PeriodicalId":243739,"journal":{"name":"2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS)","volume":"70 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114330286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Change Point Detection for Time Series Data in Complex Systems 复杂系统中时间序列数据的变化点检测

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090392

Xundong Gong, Jia Ma, Ming Chen, Shaolei Zong, Chunshan Liu

引用次数: 0

A Hybrid Method for Signal Probability Estimation with Combinational Circuits 一种组合电路信号概率估计的混合方法

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090301

Chunhong Chen, Suoyue Zhan

引用次数: 0

A 60V Input Integrated 3-to-1 Dual Inductor Hybrid Dickson Converter 60V输入集成3对1双电感混合迪克森转换器

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090294

Quyet Nguyen, Huy-Dung Han, L. Pham-Nguyen, Hanh-Phuc Le

引用次数: 0

Multi-modal Asymmetric Autoencoders for Massive Photo Collection Applications 用于大规模照片收集应用的多模态非对称自动编码器

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090330

Aymen Hamrouni, Hakim Ghazzai, Y. Massoud

{"title":"Multi-modal Asymmetric Autoencoders for Massive Photo Collection Applications","authors":"Aymen Hamrouni, Hakim Ghazzai, Y. Massoud","doi":"10.1109/APCCAS55924.2022.10090330","DOIUrl":"https://doi.org/10.1109/APCCAS55924.2022.10090330","url":null,"abstract":"There has been an abundant use of applications where many photos obtained from camera-equipped devices can be leveraged and exploited to enable emerging services, e.g., mobile crowdsourcing. These systems usually collect a large data stream of images coming from different heterogeneous sources (e.g, IoT devices and humans) in an inadvertent way. Due to the limitations and challenges related to communication bandwidth, storage, and processing capabilities, it is unwise to transfer unselectively all the photos since most of them often either contain duplicate information, are inaccurate, or are just falsely submitted. In this paper, we propose to design a smart image selection procedure using an asymmetric multi-modal neural network autoencoder to select a subset of photos that has high utility coverage for multiple incoming streams. The proposed system enables selecting high context data from an evolving picture stream and ensures relevance. The approach uses the photo's metadata such as geo-location and timestamps along with the pictures' semantics to decide which photos can be submitted and which ones must be discarded. Simulation results for two different multi-modal autoencoder architectures indicate that a mixed asymmetric stacked autoencoder approach can yield better results outperforming the concatenated input autoencoder while leveraging user-side rendering to reduce bandwidth consumption and computational overhead.","PeriodicalId":243739,"journal":{"name":"2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS)","volume":"66 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122582949","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A 0.004-mm2 O.7-V 31.654-μW BPSK Demodulator Incorporating Dual-Path Loop Self-Biased PLL 一种0.004 mm2 0.7 v 31.654 μ w双路自偏置锁相环BPSK解调器

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090331

Yixi Li, Xinyu Shen, Zhao Zhang, Guike Li, Tao Yin, Nan Qi, Jian Liu, N. Wu, Liyuan Liu, Yong Chen, Zhao Zhang

引用次数: 1

An Energy-Efficient Readout Circuit Based on Incremental Delta-Sigma ADC with Decimation Filter for CMOS Image Sensors 基于增量δ - σ ADC和抽取滤波器的CMOS图像传感器节能读出电路

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090386

Junsheng Chen, Lingxin Meng, Menglian Zhao, Z. Tan

引用次数: 0

A Dynamic JPEG CODEC with Adaptive Quantization Table for Frame Storage Compression 基于自适应量化表的动态JPEG编解码器

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090323

Yicheng Ouyang, Xianglong Wang, Gang Shi, Lei Chen, F. An

引用次数: 1

Design of SEL Self-Recovery Hardness for 90nm COTS Devices Using R-C-S Network with DC-DC Converter 基于R-C-S网络和DC-DC变换器的90nm COTS器件SEL自恢复硬度设计

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090280

Jindou Xin, Xiang Zhu, Yingqi Ma, Jianwei Han

{"title":"Design of SEL Self-Recovery Hardness for 90nm COTS Devices Using R-C-S Network with DC-DC Converter","authors":"Jindou Xin, Xiang Zhu, Yingqi Ma, Jianwei Han","doi":"10.1109/APCCAS55924.2022.10090280","DOIUrl":"https://doi.org/10.1109/APCCAS55924.2022.10090280","url":null,"abstract":"An efficient Single Event Latch-up (SEL) self-recovery hardness design using an R-C-S network in the front of a DC-DC converter is proposed in this brief. Conventional SEL hardness assurance is suffering from increased circuit design area and low reliability of the COTS devices. To overcome these problems., the proposed design makes a further contribution to the hardness of SEL. The operational mechanism of the proposed design is that the R-C-S network changes the latch-up voltage of the device through the intrinsic properties of the DC-DC converter. The latch-up voltage is carried through three phases of dropping, oscillation and rising, enabling the device to implement hardness and self-recovery. Meanwhile, simulation and laser tests are employed to demonstrate the validity of the proposed method. Additionally, through comparison with conventional hardness designs of power-off restarting and current limiting by resistor, it is evidently observed that the proposed method has current continuity characteristics and uses 5 components instead of 11 for power-off restarting. And the proposed method effectively exits the device from SEL and reduces the error bits by 72.2% compared to conventional resistor hardness.","PeriodicalId":243739,"journal":{"name":"2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132308369","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A 300-GHz Wideband Injection Locked Frequency Quadrupler in 250-nm InP DHBT Technology 250纳米InP DHBT技术中的300 ghz宽带注入锁频四倍器

2022 IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Pub Date : 2022-11-11 DOI: 10.1109/APCCAS55924.2022.10090284

Surajit Kumar Nath, Daekeun Yoon

引用次数: 0