Search

Your search keyword '"Chen, Jian-Wen"' showing total 569 results
Start Over Author "Chen, Jian-Wen"
569 results on '"Chen, Jian-Wen"'

5. RNNAccel: A Fusion Recurrent Neural Network Accelerator for Edge Intelligence

Author

Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Chiung-Liang, Chen, Pin-Han, and Lin, Youn-Long

Subjects
Computer Science - Neural and Evolutionary Computing, Computer Science - Hardware Architecture
Abstract

Many edge devices employ Recurrent Neural Networks (RNN) to enhance their product intelligence. However, the increasing computation complexity poses challenges for performance, energy efficiency and product development time. In this paper, we present an RNN deep learning accelerator, called RNNAccel, which supports Long Short-Term Memory (LSTM) network, Gated Recurrent Unit (GRU) network, and Fully Connected Layer (FC)/ Multiple-Perceptron Layer (MLP) networks. This RNN accelerator addresses (1) computing unit utilization bottleneck caused by RNN data dependency, (2) inflexible design for specific applications, (3) energy consumption dominated by memory access, (4) accuracy loss due to coefficient compression, and (5) unpredictable performance resulting from processor-accelerator integration. Our proposed RNN accelerator consists of a configurable 32-MAC array and a coefficient decompression engine. The MAC array can be scaled-up to meet throughput requirement and power budget. Its sophisticated off-line compression and simple hardware-friendly on-line decompression, called NeuCompression, reduces memory footprint up to 16x and decreases memory access power. Furthermore, for easy SOC integration, we developed a tool set for bit-accurate simulation and integration result validation. Evaluated using a keyword spotting application, the 32-MAC RNN accelerator achieves 90% MAC utilization, 1.27 TOPs/W at 40nm process, 8x compression ratio, and 90% inference accuracy., Comment: This is a paper summited in vlsicad2020 conference in Taiwan. For more information about RNNAccel, see https://neuchips.ai/rnnaccel

Published
2020

10. High Spectral Efficient Free-Space Optical Communication Using Multi-Stage Noise Shaping Delta-Sigma Modulation With Partial Transmit Sequence Algorithm and Arithmetic Coding

Author

Jian, Yin-He, Chen, Jian-Wen, Wang, Chih-Chun, Wei, Tzu-Chieh, and Chow, Chi-Wai

Abstract

We propose and experimentally demonstrate an optical wireless communication (OWC)-based mobile-fronthaul (MFH) system achieving recorded highest bit efficiencies (BEs) of 0.848 satisfying the soft-decision forward-error-correction (FEC, i.e., bit-error-rate, BER = 4 × 10-2) and 0.709 satisfying the hard-decision-FEC (i.e., BER = 3.8 × 10-3) among 1-bit DSMs. A proof-of-concept 200 m free-space transmission is demonstrated. We first propose utilizing the Hughes-Hartogs (HH) algorithm for the bit-power loading, followed by the partial transmit sequence (PTS) algorithm for reducing the peak-to-average power ratio during the orthogonal frequency division multiplexing. Then, we propose arithmetic coding (AC) in the source coding scheme to further enhance BEs. Here, we experimentally illustrate that by using the proposed HH, PTS and AC algorithms, more than 21% improvement is achieved, which is equal to an increase in the data rate of 6 Gbps per wavelength. We also experimentally illustrate that by employing wavelength division multiplexing in the proposed OWC-based MFH system, a total capacity of 1.163 Tbps (i.e., 28.57 Gbaud × 0.848 × 48 wavelengths) can be attained.

Published
2024
Full Text
View/download PDF

11. Optimized Electrophoretic Display: Dynamic Multiwaveform Invocation With Localized Time-Division for Low-Flicker Video Display

Author

Li, Shi-Xiao, Wang, Zi-Yang, Zhang, Heng, Bo, Xiao, Chen, Jian-Wen, Xu, Yi-Jian, Liu, Wei, Bai, Peng-Fei, Liu, Lin-Wei, and Zhou, Guo-Fu

Abstract

Capitalizing on their unique bistable and reflective attributes, electrophoretic displays (EPDs) provide a paper-like viewing experience with remarkably low power usage. This has driven their broad application across various fields, including eye-friendly E-readers, energy-saving electronic labels, cutting-edge wearable technology, outdoor displays, and intelligent low-power devices. In this study, a dynamic multiwaveform invocation and local time-division refresh (LTDR) technology are introduced. On the one hand, we compressed and modified traditional driving waveform to various extents first, forming optimized traditional driving waveform, fast driving waveform, and speed driving waveform. Subsequently, high-frequency oscillation waveform was incorporated into the fast driving waveform, and a variety of waveforms were dynamically invoked based on different display scenarios. On the other hand, we proposed an LTDR method, to refresh the resetting phase and driving phase in different regions simultaneously, thereby reducing the flicker effects caused by refreshing. The technology could balance display quality with refresh speed and eliminate flicker caused by traditional waveforms completely. Compared to the traditional waveform, the proposed technique can display 16-level grayscales while reducing the driving time and modulation depth by 60% and 95%, respectively, which greatly improves the reading experience of E-paper. In addition, the fast driving waveforms proposed can achieve low flicker and smooth video display in EPDs. Compared with the latest video display technology, its average structural similarity index measure (SSIM) is improved by 60.29% and energy consumption had reduced by 60.00%, which meets the demand for streaming video display on E-paper.

Published
2024
Full Text
View/download PDF

14. Taiwan

Author

Denis, Vianney, Soto, Derek, De Palmas, Stéphane, Lin, Yu Ting Vicky, Benayahu, Yehuda, Huang, Yusheng Michael, Liu, Shao-Lun, Chen, Jian-Wen, Chen, Qi, Sturaro, Nicolas, Ho, Ming-Jay, Su, Yeng, Dai, Chang Feng, Chen, Chaolun Allen, Riegl, Bernhard M., Series Editor, Dodge, Richard E., Series Editor, Loya, Yossi, editor, Puglise, Kimberly A., editor, and Bridge, Tom C.L., editor

Published
2019
Full Text
View/download PDF

16. Preliminary results of a phase II study of surufatinib plus sintilimab, nab-paclitaxel and gemcitabine (AG) as first-line therapy in patients (pts) with locally advanced or metastatic pancreatic adenocarcinoma (mPDAC).

Author

Zhang, Dong-Sheng, primary, Liu, Fu-Rong, additional, Lu, Yun-Xin, additional, Bai, Bing, additional, Zhang, Yang, additional, Wang, Zhi-Qiang, additional, Liu, Yu, additional, Chen, Jian-Wen, additional, Wang, Feng, additional, and Xu, Rui-Hua, additional

Published
2024
Full Text
View/download PDF

36. Design of a Passive Silicon-on-Insulator-Based On-Chip Optical Circulating Network Supporting Mode Conversion and High Optical Isolation.

Author

Lin, Yuan-Zeng, Chen, Jian-Wen, Chow, Chi-Wai, and Yeh, Chien-Hung

Subjects
PASSIVE optical networks, SOCIAL networks, INTEGRATED circuits, INTERNET traffic, OPTICAL control, PHOTONICS, INTEGRATED optics, BANDWIDTHS
Abstract

Over the past few decades, on-chip photonic integrated circuits based on silicon photonics (SiPh) platforms have gained widespread attention due to the fact that they offer many advantages, such as high bandwidth, low loss, compact size, low power consumption, and high integration with different photonic devices. The demand for high-speed and high-performance SiPh devices is driven by the significant increase in demand for Internet traffic. In photonic integrated circuits, controlling optical signals to make them circulate in a specific direction is a highly researched area of study. However, achieving a purely passive on-chip optical circulating network on a SiPh platform is very challenging. Therefore, we propose and demonstrate, through simulations, an on-chip optical circulator network on a silicon-on-insulator (SOI) platform. The proposed device can also support mode conversion. The proposed on-chip optical circulating network consists of two kinds of tailor-made multi-mode interferometer (MMI) structures and waveguide crossings. Through the optical power division and mode combination capabilities of the MMI, an optical circulating network supporting high optical isolation and mode conversion is achieved. The proposed optical circulating network has a loss of 1.5 dB at each output port, while maintaining a high isolation of 35 dB in the transmission window from 1530 nm to 1570 nm. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

40. CABAC Encoder

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

41. Fractional Motion Estimation

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

42. Motion Compensation

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

43. System Integration

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

44. Transform Coding

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

45. Deblocking Filter

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

46. Integer Motion Estimation

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

48. Intra Prediction

Author

Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Huang-Chih, Chen, Jian-Wen, Lin, Youn-Long Steve, Kao, Chao-Yang, Kuo, Hung-Chih, and Chen, Jian-Wen

Published
2010
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results