Hui Zhang

Hui Zhang

Auto-generated profile, based on publicly available information

Hui Zhang

Institute of Precision Optical Engineering, School of Physics Science and Engineering, Tongji University, Shanghai, China

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
2 December 2022 in Science Advances
Verified
Journal Article
Open access
Yuzhi ShiYuzhi Shi
Yuzhi Shi
TZTongtong Zhu
Tongtong Zhu
JLJing-Quan Liu
Jing-Quan Liu
DTDin Ping Tsai
Din Ping Tsai
HZHui Zhang
Hui Zhang
Ai Qun LiuAi Qun Liu
Ai Qun Liu
Transverse spin momentum related to the spin angular momentum (SAM) of light has been theoretically studied recently and predicted to generate an intriguing optical lateral force (OLF). Despite extensive studies, there is no direct experimental evidence of a stable OLF resulting from the dominant SAM rather than the ubiquitous spin-orbit interaction in a single light beam. Here, we theoretically unveil the nontrivial physics of SAM-correlated OLF, showing that the SAM is a dominant factor for the OLF on a nonabsorbing particle, while an additional force from the canonical (orbital) momentum is exhibited on an absorbing particle due to the spin-orbit interaction. Experimental results demonstrate the bidirectional movement of 5-μm-diameter particles on both sides of the beam with opposite spin momenta. The amplitude and sign of this force strongly depend on the polarization. Our optofluidic platform advances the exploitation of exotic forces in systems with a dominant SAM, facilitating the exploration of fascinating light-matter interactions.

7 October 2022 in Science Advances
Verified
Journal Article
Open access
HZHui Zhang
Hui Zhang
LWLingxiao Wan
Lingxiao Wan
THTobias Haug
Tobias Haug
WMWai-Keong Mok
Wai-Keong Mok
SPStefano Paesani
Stefano Paesani
Ai Qun LiuAi Qun Liu
Ai Qun Liu
Quantum autoencoders serve as efficient means for quantum data compression. Here, we propose and demonstrate their use to reduce resource costs for quantum teleportation of subspaces in high-dimensional systems. We use a quantum autoencoder in a compress-teleport-decompress manner and report the first demonstration with qutrits using an integrated photonic platform for future scalability. The key strategy is to compress the dimensionality of input states by erasing redundant information and recover the initial states after chip-to-chip teleportation. Unsupervised machine learning is applied to train the on-chip autoencoder, enabling the compression and teleportation of any state from a high-dimensional subspace. Unknown states are decompressed at a high fidelity (~0.971), obtaining a total teleportation fidelity of ~0.894. Subspace encodings hold great potential as they support enhanced noise robustness and increased coherence. Laying the groundwork for machine learning techniques in quantum systems, our scheme opens previously unidentified paths toward high-dimensional quantum computing and networking.

26 November 2024 in Nature Communications
Verified
Journal Article
Open access
PDF file
JZJing Zhou
Jing Zhou
HZHui Zhang
Hui Zhang
QQQifeng Qiao
Qifeng Qiao
HCHeng Chen
Heng Chen
QHQian Huang
Qian Huang
Chengkuo LeeChengkuo Lee
Chengkuo Lee
Silicon photonics enables the construction of chip-scale spectrometers, in which those using a single tunable interferometer provide a simple and cost-effective solution. Among various tuning mechanisms, electrostatic MEMS reconfiguration stands out as an ideal candidate, given its high tuning efficiency and ultra-low power consumption. Nonetheless, MEMS devices face significant noise challenges arising from their susceptible minuscule components, adversely impacting spectral resolution. Here, we propose a distinct paradigm of spectrometers through synergizing an easily-fabricated MEMS-reconfigurable low-loss waveguide coupler on a silicon photonic chip and a convolutional autoencoder denoising (CAED) mechanism. The spectrometer offers a 300 nm bandwidth and a reconstruction resolution of 0.3 nm in a noise-free condition. In a noisy environment with a signal-to-noise ratio as low as 30 dB, the reconstruction resolution of the interferograms processed by the CAED exhibits an enhancement from 1.2 to 0.4 nm, approaching the noise-free value. Our technology is envisaged to provide a powerful and cost-effective solution for applications requiring accurate, broadband, and energy-efficient spectral analysis.

27 September 2023 in PRX Quantum
Verified
Journal Article
Open access
PDF file
HZHui Zhang
Hui Zhang
LWLingxiao Wan
Lingxiao Wan
SPStefano Paesani
Stefano Paesani
ALAnthony Laing
Anthony Laing
YSYuzhi Shi
Yuzhi Shi
Ai Qun LiuAi Qun Liu
Ai Qun Liu
Integrated photonics provides a versatile platform for encoding and processing quantum information. However, the encoded quantum states are sensitive to noise, which limits their capability to perform complicated quantum computations. Here, we use a five-qubit linear cluster state on a silicon photonic chip to implement a quantum error-correction code and demonstrate its capability of identifying and correcting a single-qubit error. The encoded quantum information is reconstructed from a single-qubit error and an average state fidelity of 0.863±0.032 is achieved for different input states. We further extend the scheme to demonstrate a fault-tolerant measurement-based quantum computation (MBQC) on stabilizer formalism that allows us to redo the qubit operation against the failure of the teleportation process. Our work provides a proof-of-concept working prototype of error correction and MBQC in an integrated photonic chip.

25 December 2022 in Laser & Photonics Reviews
Verified
Journal Article
Open access
HZHui Zhang
Hui Zhang
JLJonathan Wei Zhong Lau
Jonathan Wei Zhong Lau
LWLingxiao Wan
Lingxiao Wan
LSLiang Shi
Liang Shi
YSYuzhi Shi
Yuzhi Shi
ALAi Qun Liu
Ai Qun Liu
Machine learning methods have revolutionized the discovery process of new molecules and materials. However, the intensive training process of neural networks for molecules with ever‐increasing complexity has resulted in exponential growth in computation cost, leading to long simulation time and high energy consumption. Photonic chip technology offers an alternative platform for implementing neural networks with faster data processing and lower energy usage compared to digital computers. Photonics technology is naturally capable of implementing complex‐valued neural networks at no additional hardware cost. Here, the capability of photonic neural networks for predicting the quantum mechanical properties of molecules is demonstrated. To the best of knowledge, this work is the first to harness photonic technology for machine learning applications in computational chemistry and molecular sciences, such as drug discovery and materials design. It is further shown that multiple properties can be learned simultaneously in a photonic chip via a multi‐task regression learning algorithm, which is also the first of its kind as well, as most previous works focus on implementing a network in the classification task.

23 February 2024 in Laser & Photonics Reviews
Verified
Journal Article
Open access
YZYuancheng Zhan
Yuancheng Zhan
HZHui Zhang
Hui Zhang
HLHexiang Lin
Hexiang Lin
LCLip Ket Chin
Lip Ket Chin
HCHong Cai
Hong Cai
ALAi Qun Liu
Ai Qun Liu
Photonic neural networks (PNNs) have emerged as promising alternatives to traditional electronic neural networks. However, the training of PNNs, especially the chip implementation of analytic gradient descent algorithms that are recognized as highly efficient in traditional practice, remains a major challenge because physical systems are not differentiable. Although training methods such as gradient‐free and numerical gradient methods are proposed, they suffer from excessive measurements and limited scalability. State‐of‐the‐art in situ training method is also cost‐challenged, requiring expensive in‐line monitors and frequent optical I/O switching. Here, a physics‐aware analytic‐gradient training (PAGT) method is proposed that calculates the analytic gradient in a divide‐and‐conquer strategy, overcoming the difficulty induced by chip non‐differentiability in the training of PNNs. Multiple training cases, especially a generative adversarial network, are implemented on‐chip, achieving a significant reduction in time consumption (from 31 h to 62 min) and a fourfold reduction in energy consumption, compared to the in situ method. The results provide low‐cost, practical, and accelerated solutions for training hybrid photonic‐digital electronic neural networks.