arXiv subjects

Jia-Nan Wu

Publications and source records attributed to Jia-Nan Wu.

4 recordsLinked to original sources

arXiv2026

Dynamically preparing robust Bell states by time-boundary engineering

Quantum entanglement is essential for modern quantum information processing. Entanglement gates convert initially non-entangled states into entangled ones by applying time-dependent parametric pulses. While Bell state preparation has been experimentally validated in various platforms, its stability and fidelity are constrained by environmental decoherence and parametric fluctuations.Here, we propose a dynamical framework for preparing robust Bell states by leveraging time-boundary engineering and momentum-space projective measurements within Su-Schrieffer-Heeger (SSH) systems. Employing Lindblad master equation, we theoretically demonstrate that the prepared Bell states exhibit remarkable robustness against both environmental decoherence and parametric time fluctuations, achieving a nearly perfect quantum fidelity, with momentum conservation law governing this robust behavior. To enrich Bell states in momentum space, multi-band SSH models are designed to induce multifold time scattering processes. This time-boundary engineering framework is applicable to both fermionic and bosonic excitations, offering a robust paradigm for generating Bell states in quantum communication and quantum computation.

quant-ph↗

arXiv2023

Statistics of tens-of-photon states scattered by optical cavity, two-level atom and Jaynes-Cummings emitter

Manipulating photon states serves as a primary requirement for various optical devices and is of high relevance for quantum information technology. Nevertheless, the fundamental theoretical framework for tens-of-photon states has not been established. This study successfully establishes the matrix-product-state theory to explore the statistics of the tens-of-photon states scattered by optical cavities (OCs), two-level atoms (TLAs), and Jaynes-Cummings emitters (JCEs) in waveguide-QED systems. Taking 10-photon states as an example, we reveal some novel physical results that differ from those for few-photon cases. We verify that OCs do not change the statistics of the incident photon states, being independent of the photon number. However, for the TLAs and JCEs, the photon number strongly impacts the photon bunching and anti-bunching behaviors. As the photon number increases, there exists a maximum strength for the photon-photon correlation induced by the JCE. Especially, the scattered waves by the TLA (or JCE) exhibit extremely different statistics behaviors for the 10-photon cases from those for the bi-photon. These distinguishable conclusions for the tens-of-photon states and the matrix-product-state theory pave the way for the multi-photon manipulation.

physics.optics↗

arXiv2016

Exploiting Local Structures with the Kronecker Layer in Convolutional Networks

In this paper, we propose and study a technique to reduce the number of parameters and computation time in convolutional neural networks. We use Kronecker product to exploit the local structures within convolution and fully-connected layers, by replacing the large weight matrices by combinations of multiple Kronecker products of smaller matrices. Just as the Kronecker product is a generalization of the outer product from vectors to matrices, our method is a generalization of the low rank approximation method for convolution neural networks. We also introduce combinations of different shapes of Kronecker product to increase modeling capacity. Experiments on SVHN, scene text recognition and ImageNet dataset demonstrate that we can achieve $3.3 \times$ speedup or $3.6 \times$ parameter reduction with less than 1\% drop in accuracy, showing the effectiveness and efficiency of our method. Moreover, the computation efficiency of Kronecker layer makes using larger feature map possible, which in turn enables us to outperform the previous state-of-the-art on both SVHN(digit recognition) and CASIA-HWDB (handwritten Chinese character recognition) datasets.

cs.CV↗

arXiv2015

Compression of Fully-Connected Layer in Neural Network by Kronecker Product

In this paper we propose and study a technique to reduce the number of parameters and computation time in fully-connected layers of neural networks using Kronecker product, at a mild cost of the prediction quality. The technique proceeds by replacing Fully-Connected layers with so-called Kronecker Fully-Connected layers, where the weight matrices of the FC layers are approximated by linear combinations of multiple Kronecker products of smaller matrices. In particular, given a model trained on SVHN dataset, we are able to construct a new KFC model with 73\% reduction in total number of parameters, while the error only rises mildly. In contrast, using low-rank method can only achieve 35\% reduction in total number of parameters given similar quality degradation allowance. If we only compare the KFC layer with its counterpart fully-connected layer, the reduction in the number of parameters exceeds 99\%. The amount of computation is also reduced as we replace matrix product of the large matrices in FC layers with matrix products of a few smaller matrices in KFC layers. Further experiments on MNIST, SVHN and some Chinese Character recognition models also demonstrate effectiveness of our technique.

cs.NE↗