Zhe Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zhe Wang has a total of 32 co-authors. Here are the top co-authors by number of shared publications:
- Jian Liu
Chinese Academy of Sciences, State Key Laboratory of Semiconductor Physics and Chip Technologies, Institute of Semiconductors, Beijing, China
6 shared publications - Liyuan Liu
Institute of Semiconductors, Chinese Academy of Sciences, Beijing, China
6 shared publications - Nanjian Wu
Chinese Academy of Sciences, State Key Laboratory of Semiconductor Physics and Chip Technologies, Institute of Semiconductors, Beijing, China
5 shared publications - Runjiang Dou
Chinese Academy of Sciences
4 shared publications - Shuangming Yu
Chinese Academy of Sciences
4 shared publications - Xu Yang
Institute of Semiconductors, Chinese Academy of Sciences
4 shared publications
Latest Publications
- 8 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 61 (3) , 1164-1180
A 10 000-Inference/s Bio-Inspired Spiking Vision Chip Based on an End-to-End SNN Embedding Image Signal Enhancement
Edge machine vision requires a vision chip to perform imaging and visual recognition with high speed, high accuracy, and compact size. Current vision chips adopt a two-stage workflow, which uses computer vision processing and a deep neural network (DNN) to perform signal enhancement and recognition, respectively. However, the two stages involve different operations and are separately optimized, which causes degraded recognition accuracy, low inference rate, and huge hardware overhead. To address it, we propose a vision chip that integrates a single-photon avalanche diode (SPAD) image sensor and a reconfigurable spiking neural network (SNN) processor. Inspired by the human visual system that employs a full SNN for signal enhancement and recognition, the vision chip acquires spike-based images and performs an end-to-end SNN embedding image signal enhancement (ISE) for object recognition. The ISE and recognition are trained jointly in a unified deep learning framework to increase recognition accuracy and inference speed. The vision chip only integrates an SNN processor to perform ISE and recognition by being reconfigured into the low-precision multiply-accumulation (MAC) mode or accumulation (ACC) mode, respectively, thereby avoiding a dedicated image signal processor (ISP). Based on a 55-nm CMOS technology, the fabricated vision chip integrates $128{\times }128$ pixels and can implement 4096 ACC/MAC components. The chip was measured to exhibit: 1) a 99.2% accuracy with 10000 inferences/s at 10 lx and 95.3% at 1.43 mlx on the MNIST dataset and 2) a 94.32% accuracy with color mosaic images on the ETH-80 dataset. Our chip maintains a relatively high recognition performance even under an illumination one order of magnitude lower than in prior vision chips.
- 8 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 61 (3) , 1164-1180
A 10 000-Inference/s Bio-Inspired Spiking Vision Chip Based on an End-to-End SNN Embedding Image Signal Enhancement
Edge machine vision requires a vision chip to perform imaging and visual recognition with high speed, high accuracy, and compact size. Current vision chips adopt a two-stage workflow, which uses computer vision processing and a deep neural network (DNN) to perform signal enhancement and recognition, respectively. However, the two stages involve different operations and are separately optimized, which causes degraded recognition accuracy, low inference rate, and huge hardware overhead. To address it, we propose a vision chip that integrates a single-photon avalanche diode (SPAD) image sensor and a reconfigurable spiking neural network (SNN) processor. Inspired by the human visual system that employs a full SNN for signal enhancement and recognition, the vision chip acquires spike-based images and performs an end-to-end SNN embedding image signal enhancement (ISE) for object recognition. The ISE and recognition are trained jointly in a unified deep learning framework to increase recognition accuracy and inference speed. The vision chip only integrates an SNN processor to perform ISE and recognition by being reconfigured into the low-precision multiply-accumulation (MAC) mode or accumulation (ACC) mode, respectively, thereby avoiding a dedicated image signal processor (ISP). Based on a 55-nm CMOS technology, the fabricated vision chip integrates $128{\times }128$ pixels and can implement 4096 ACC/MAC components. The chip was measured to exhibit: 1) a 99.2% accuracy with 10000 inferences/s at 10 lx and 95.3% at 1.43 mlx on the MNIST dataset and 2) a 94.32% accuracy with color mosaic images on the ETH-80 dataset. Our chip maintains a relatively high recognition performance even under an illumination one order of magnitude lower than in prior vision chips.
- 18 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (11) , 6652-6664
A 75.6 Gb/s 22-bit Floating-Point Coarse-Grained Versatile DSP Embedding 26K-Point Baseband Signal Processing and 2048 × 256-Point Complex FFT
As millimeter-wave radar technology advances, modern domain-specific digital signal processors (DSPs) struggle to balance versatility and processing scale, while suffering from low data precision and data throughput. To address these challenges, this paper presents a novel coarse-grained versatile DSP (CVDSP). The CVDSP introduces an architecture based on multi-level finite state machines and a custom instruction set to support various algorithms through flexible dataflow. The efficient large-scale PEs are designed with 22-bit floating-point precision to handle 26K-point baseband signal processing and $2048\times 256$ -point complex fast Fourier transform. Cooperating with a high-bandwidth instruction-free memory access network, the CVDSP achieves relatively high data throughput. Fabricated in a 65-nm CMOS process, experimental results show that the peak energy efficiency and data throughput of the CVDSP are 299.5 GMACs/J and 75.6 Gb/s. The CVDSP demonstrates fully on-chip implementation of the FMCW radar, Pulse-Doppler radar, and spectrometer algorithms.
- 18 June 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Circuits and Systems I: Regular Papers
- Vol. 72 (11) , 6652-6664
A 75.6 Gb/s 22-bit Floating-Point Coarse-Grained Versatile DSP Embedding 26K-Point Baseband Signal Processing and 2048 × 256-Point Complex FFT
As millimeter-wave radar technology advances, modern domain-specific digital signal processors (DSPs) struggle to balance versatility and processing scale, while suffering from low data precision and data throughput. To address these challenges, this paper presents a novel coarse-grained versatile DSP (CVDSP). The CVDSP introduces an architecture based on multi-level finite state machines and a custom instruction set to support various algorithms through flexible dataflow. The efficient large-scale PEs are designed with 22-bit floating-point precision to handle 26K-point baseband signal processing and $2048\times 256$ -point complex fast Fourier transform. Cooperating with a high-bandwidth instruction-free memory access network, the CVDSP achieves relatively high data throughput. Fabricated in a 65-nm CMOS process, experimental results show that the peak energy efficiency and data throughput of the CVDSP are 299.5 GMACs/J and 75.6 Gb/s. The CVDSP demonstrates fully on-chip implementation of the FMCW radar, Pulse-Doppler radar, and spectrometer algorithms.
- 10 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (9) , 2934-2947
A 3 THz CMOS Image Sensor
This article presents a 3 THz CMOS image sensor (Tera-CIS). The sensor has a column-parallel readout (CPRO) architecture that integrates an antenna-coupled pixel array and CPRO circuit chains on a monolithic chip. The proposed compact two-transistor (2T) pixel adopts a step-covered patch antenna and a defected ground structure (DGS) to obtain sufficient sensitivity and bandwidth. The step-covered patch antenna model is developed to predict pixel performances precisely. The DGS structure suppresses mutual coupling among adjacent pixels and shrinks the pixel pitch. In the CPRO circuit chain, chopping and oversampling techniques are employed to reduce noise and flexibly balance the dynamic range (DR) characteristics with the imaging rate. A digital decimation filter (DDF) with a time-multiplexing fashion is adopted to alleviate resource pressure. A 16.4k-pixel Tera-CIS was fabricated with a standard $0.18~\mu \text{m}$ CMOS process. A 3 THz imaging platform with four different continuous-wave (CW) terahertz quantum cascade lasers was established. The pixel sensitivity was 753 V/W at 3.4 THz, with a measured detection bandwidth of 0.78 THz (from 3.08 to 3.86 THz). The DR in the voltage domain of the sensor reached 73 dB at 8 fps while the maximum DR in the power domain was 39.8 dB. Meanwhile, the sensor can operate up to 130 fps. The imaging system can achieve high-resolution imaging and clearly identify concealed objects.
- 10 April 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Solid-State Circuits
- Vol. 59 (9) , 2934-2947
A 3 THz CMOS Image Sensor
This article presents a 3 THz CMOS image sensor (Tera-CIS). The sensor has a column-parallel readout (CPRO) architecture that integrates an antenna-coupled pixel array and CPRO circuit chains on a monolithic chip. The proposed compact two-transistor (2T) pixel adopts a step-covered patch antenna and a defected ground structure (DGS) to obtain sufficient sensitivity and bandwidth. The step-covered patch antenna model is developed to predict pixel performances precisely. The DGS structure suppresses mutual coupling among adjacent pixels and shrinks the pixel pitch. In the CPRO circuit chain, chopping and oversampling techniques are employed to reduce noise and flexibly balance the dynamic range (DR) characteristics with the imaging rate. A digital decimation filter (DDF) with a time-multiplexing fashion is adopted to alleviate resource pressure. A 16.4k-pixel Tera-CIS was fabricated with a standard $0.18~\mu \text{m}$ CMOS process. A 3 THz imaging platform with four different continuous-wave (CW) terahertz quantum cascade lasers was established. The pixel sensitivity was 753 V/W at 3.4 THz, with a measured detection bandwidth of 0.78 THz (from 3.08 to 3.86 THz). The DR in the voltage domain of the sensor reached 73 dB at 8 fps while the maximum DR in the power domain was 39.8 dB. Meanwhile, the sensor can operate up to 130 fps. The imaging system can achieve high-resolution imaging and clearly identify concealed objects.
- Open Access
- 9 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Journal
- Vol. 15 (6) , 1-9
Modeling and Analysis of Noise Reduction Method in SPAD-Based LiDAR System
Ambient light interference stands as the primary source of noise encountered by the light detection and ranging (LiDAR) systems based on single-photon avalanche diode (SPAD), which seriously restricts the detection range and accuracy of sensors. Multi-event mode and time-gating method are extensively employed due to their effective suppression ability of the ambient light interference with easily implemented circuit structure. However, the current research on these two operational modes is limited to qualitative description, lacking a comprehensive theoretical analysis. This paper investigates the impact of the multi-event mode and time-gating method on enhancing the imaging capabilities of the sensor, employing a probability distribution perspective. A behavioral model is established, enabling the simulation of the comprehensive functionality of a SPAD-based LiDAR system in either first-hit mode or multi-event mode, including the histogramming, coincident detection, time- gating functions. The model can serve as a substitute for the actual circuit in system function validation and parameter optimization, thereby significantly reducing the design and simulation cycle.
- 9 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Photonics Journal
- Vol. 15 (6) , 1-9
Open AccessModeling and Analysis of Noise Reduction Method in SPAD-Based LiDAR System
Ambient light interference stands as the primary source of noise encountered by the light detection and ranging (LiDAR) systems based on single-photon avalanche diode (SPAD), which seriously restricts the detection range and accuracy of sensors. Multi-event mode and time-gating method are extensively employed due to their effective suppression ability of the ambient light interference with easily implemented circuit structure. However, the current research on these two operational modes is limited to qualitative description, lacking a comprehensive theoretical analysis. This paper investigates the impact of the multi-event mode and time-gating method on enhancing the imaging capabilities of the sensor, employing a probability distribution perspective. A behavioral model is established, enabling the simulation of the comprehensive functionality of a SPAD-based LiDAR system in either first-hit mode or multi-event mode, including the histogramming, coincident detection, time- gating functions. The model can serve as a substitute for the actual circuit in system function validation and parameter optimization, thereby significantly reducing the design and simulation cycle.
- Open Access
- 30 June 2020
- Published by China Science Publishing & Media Ltd. in Terahertz Science and Technology
- Vol. 13 (2) , 61-72
High performance terahertz quantum cascade lasers
Terahertz region is the electromagnetic gap between the infrared optoelectronics and the high frequency electronics, which is of broad prospects in applications. The application requirements drive the rapid development in Terahertz technologies including sources, detectors and systems. In the last two decades, quantum cascade laser has made great progress as one of the most promising terahertz sources. In this paper, we present the development of terahertz quantum cascade lasers in our group.
- 30 June 2020
- Published by China Science Publishing & Media Ltd. in Terahertz Science and Technology
- Vol. 13 (2) , 61-72
Open AccessHigh performance terahertz quantum cascade lasers
Terahertz region is the electromagnetic gap between the infrared optoelectronics and the high frequency electronics, which is of broad prospects in applications. The application requirements drive the rapid development in Terahertz technologies including sources, detectors and systems. In the last two decades, quantum cascade laser has made great progress as one of the most promising terahertz sources. In this paper, we present the development of terahertz quantum cascade lasers in our group.