Browsing by Author "Lee, Yu-Chi"

Now showing 1 - 5 of 5

A 2.17-mW Acoustic DSP Processor With CNN-FFT Accelerators for Intelligent Hearing Assistive Devices
(2020-08-01)
Lee, Yu-Chi
;
Chi, Tai-Shih
;
Yang, Chia-Hsiang
;
電機工程學系
;
Department of Electrical and Computer Engineering
This article presents an acoustic DSP processor containing a neural network core for intelligent hearing assistive devices. The processor includes the accelerators for convolutional neural networks (CNNs) and fast Fourier transform (FFT). The CNN-based speech enhancement algorithm predicts the desired mask for the Fourier spectrogram of the speech signal to enhance speech intelligibility. Several design techniques are applied to enable efficient hardware mapping. The computational complexity for the CNN can be reduced by 23.6% by frame sharing, and a fast mask generation + partial sums pre-computation technique further reduces output latency by up to 64%. The size of the memory for the model is reduced by 75% using weight quantization. FFT is implemented by leveraging the packing algorithm to reduce the computational complexity by 43%. Reconfigurable processing elements are shared to support both FFT and CNN, realizing a saving in the area of 42%. In addition, input sharing and output sharing are used to, respectively, reduce data movements by 94% and 75%. A reordered FFT structure also eliminates up to 256 multiplexers. Fabricated in a 40-nm CMOS technology, the chip's core area is 4.2 mm(2) and the power dissipation is 2.17 mW at a clock frequency of 5 MHz from a 0.6-V supply. The embedded CNN accelerator supports both convolutional and fully connected (FC) layers and achieves a comparable energy efficiency with state-of-the-art CNN accelerators, despite the flexibility for FFT. The speech intelligibility is enhanced by up to 41% in the low SNR regime.
A 2.17mW Acoustic DSP Processor with CNN-FFT Accelerators for Intelligent Hearing Aided Devices
(2019-01-01)
Lee, Yu-Chi
;
Chi, Tai-Shih
;
Yang, Chia-Hsiang
;
電機工程學系
;
Department of Electrical and Computer Engineering
This paper proposes an acoustic DSP processor with a neural network core for speech enhancement. Accelerators for convolutional neural network (CNN) and fast Fourier transform (FFT) are embedded. The CNN-based speech enhancement algorithm takes the speech signals spectrogram as the model's input, and predicts the desired mask of speech to enhance speech intelligibility after passing through the CNN model. An array of multiply-accumulator (MAC) and coordinate rotation digital computer (CORDIC) engines are deployed to efficiently compute linear and nonlinear functions. Hardware sharing is applied to reduce hardware area by leveraging the high similarity between CNN and FFT computations. The proposed DSP processor chip is fabricated in a 40-nm CMOS technology with a core area of 4.3 mm(2). The chip's power dissipation is 2.17 mW at an operating frequency of 5 MHz. The CNN accelerator supports both convolutional and fully-connected layers and achieves an energy efficiency of 1200-to-2180 GOPS/W, despite the flexibility for FFT. The speech intelligibility can be enhanced by up to 41% under low SNR conditions.
A 98.6 mu W Acoustic Signal Processor for Fully-Implantable Cochlear Implants
(2016)
Liu, Hao-Min
;
Lin, Yung-Jen
;
Lee, Yu-Chi
;
Lee, Cheng-Yen
;
Yang, Chia-Hsiang
This paper presents a low-power acoustic signal processor for fully-implantable cochlear implants. The developed processor supports adaptive beamforming, frequency-domain analysis, envelope detection, channel combination, and magnitude compression. Power and area are minimized by leveraging dedicated real-valued FFT, register count minimization, data allocation optimization, hardware complexity reduction, and minimum-energy-point operation. Compared to complexvalued FFT, real-valued FFT achieves 44.36% power reduction. Register count minimization and data allocation for FFT output reordering yields 28.07% and 27.09% area and power reduction, respectively. Envelope detection and log-compression are realized by hardware-efficient CORDIC engines. The processor is scalable to support various numbers of channels. This chip is implemented in 90-nm CMOS and the core area is 0.47 mm(2). It dissipates 98.6 mu W at 50 kHz, 0.33 V for a latency of 3 ms.
A Bone-Guided Cochlear Implant CMOS Microsystem Preserving Acoustic Hearing
(2017-01-01)
Qian, Xin-Hong
;
Wu, Yi-Chung
;
Yang, Tzu-Yi
;
Cheng, Cheng-Hsiang
;
Chu, Hsing-Chien
In this paper, a bone-guided cochlear implant (BGCI) SOC microsystem is proposed and designed. The BGCI microsystem uses four or more electrodes placed on the bone surface of the cochlea and one on the round window to preserve partially the acoustic hearing. The external SOC of the BGCI processes the acoustic signals and generates stimulation patterns and command that are transmitted to the implanted chip through the 13.56MHz wireless power and bidirectional data telemetry. The implanted chip with 4-channel high-voltage-tolerant stimulator generates biphasic stimulation currents up to 800 mu A and sends back the power indication and electrode impedance error signals to the external SOC. Electrical tests on the fabricated BGCI has been performed. The in-vivo animal tests on guinea pigs has shown the evoked Wave III of EABR waveforms. The proposed BGCI can be applied to patients with high-frequency hearing loss, tinnitus, and dizziness.
Design and In Vivo Verification of a CMOS Bone-Guided Cochlear Implant Microsystem
(2019-11-01)
Qian, Xin-Hong
;
Wu, Yi-Chung
;
Yang, Tzu-Yi
;
Cheng, Cheng-Hsiang
;
Chu, Hsing-Chien
Objective: To develop and verify a CMOS bone-guided cochlear implant (BGCI) microsystem with electrodes placed on the bone surface of the cochlea and the outside of round window for treating high-frequency hearing loss. Methods: The BGCI microsystem consists of an external unit and an implanted unit. The external system-on-chip is designed to process acoustic signals through an acquisition circuit and an acoustic DSP processor to generate stimulation patterns and commands that are transmitted to the implanted unit through a 13.56 MHz wireless power and bidirectional data telemetry. In the wireless power telemetry, a voltage doubler/tripler (2X/3X) active rectifier is used to enhance the power conversion efficiency and generate 2 and 3 V output voltages. In the wireless data telemetry, phase-locked loop based binary phase-shift keying and load-shift keying modulators/demodulators are adopted for the downlink and uplink data through high-Q coils, respectively. The implanted chip with four-channel high-voltage-tolerant stimulator generates biphasic stimulation currents up to 800 mu A. Results: Electrical tests on the fabricated BGCI microsystem have been performed to verify the chip functions. The in vivo animal tests in guinea pigs have shown the evoked third wave of electrically evoked auditory brainstem response waveforms. It is verified that auditory nerves can be successfully stimulated and acoustic hearing can be partially preserved. Conclusion and Significance: Different from traditional cochlear implants, the proposed BGCI microsystem is less invasive, preserves partially acoustic hearing, and provides an effective alternative for treating high-frequency hearing loss.