A5100783219- Quantum Information and Cryptography
- Quantum Computing Algorithms and Architecture
- Quantum Mechanics and Applications
- Advanced Neural Network Applications
- Radiomics and Machine Learning in Medical Imaging
- COVID-19 diagnosis using AI
- AI in cancer detection
- Domain Adaptation and Few-Shot Learning
- Multimodal Machine Learning Applications
- Advanced Image and Video Retrieval Techniques
- Microwave Engineering and Waveguides
- Topic Modeling
- Protein Structure and Dynamics
- Millimeter-Wave Propagation and Modeling
- Medical Image Segmentation Techniques
- Image Retrieval and Classification Techniques
- Antenna Design and Analysis
- Quantum optics and atomic interactions
- Photonic and Optical Devices
- Machine Learning in Healthcare
- Remote-Sensing Image Classification
- Medical Imaging and Analysis
- Video Surveillance and Tracking Methods
- Brain Tumor Detection and Classification
- Computational Drug Discovery Methods
University of Hong Kong
2021-2025
Harvard University
2024-2025
Boston University
2025
Tsinghua University
2024-2025
Affiliated Eye Hospital of Wenzhou Medical College
2024
Wenzhou Medical University
2024
Kunming University of Science and Technology
2024
Beijing Academy of Artificial Intelligence
2024
Chinese University of Hong Kong
2021-2024
Harbin Institute of Technology
2024
We analyzed the security of multiparty quantum secret sharing (MQSS) protocol recently proposed by Zhang, Li, and Man [Phys. Rev. A 71, 044301 (2005)] found that this is secure for any other eavesdropper except agent Bob who prepares signals as he can attack communication with a Trojan horse. That is, replaces single-photon signal multiphoton one Charlie cannot find cheating she does not measure photons before they run back from boss Alice, which reveals MQSS Bob. Finally, we present...
We analyzed the security of secure direct communication protocol based on secret transmitting order particles recently proposed by Zhu, Xia, Fan, and Zhang[Phys. Rev. A 73, 022338 (2006)] found that this scheme is insecure if an eavesdropper, say Eve, wants to steal message with Trojan horse attack strategies. The vital loophole in two authorized users check their quantum channel only once. Eve can insert another spy photon, invisible or a delay one each photon which sender Alice sends...
A scheme for multiparty quantum state sharing of an arbitrary two-particle is presented with Einstein-Podolsky-Rosen pairs. Any one the $N$ agents has access to regenerate original two local unitary operations if he collaborates other agents, say controllers. Moreover, each controllers required take only a product measurement ${\ensuremath{\sigma}}_{x}\ensuremath{\bigotimes}{\ensuremath{\sigma}}_{x}$ on his particles, which makes this more convenient in applications network than others. As...
We present a way for symmetric multiparty-controlled teleportation of an arbitrary two-particle entangled state based on Bell-basis measurements by using two Greenberger-Horne-Zeilinger states, i.e., sender transmits to distant receiver, one the $n+1$ agents, via control others in network. It will be shown that outcomes cases $n$ is odd or even are different principle as receiver has perform controlled-NOT operation his particles reconstructing original addition some local unitary operations...
We present two efficient quantum key distribution schemes over different collective-noise channels. The accepted hypothesis of collective noise is that photons travel inside a time window small compared to the variation noise. Noiseless subspaces are made up Bell states and spatial degree freedom introduced form nonorthogonal bases. Although these protocols resort entangled for encoding bit, receiver only required perform single-particle product measurements there no basis mismatch....
Epidermal growth factor receptor (EGFR) genotyping is critical for treatment guidelines such as the use of tyrosine kinase inhibitors in lung adenocarcinoma. Conventional identification EGFR genotype requires biopsy and sequence testing which invasive may suffer from difficulty accessing tissue samples. Here, we propose a deep learning model to predict mutation status adenocarcinoma using non-invasive computed tomography (CT). We retrospectively collected data 844 patients with pre-operative...
We present a way for entanglement purification based on two parametric down-conversion (PDC) sources with cross-Kerr nonlinearities. It is comprised of processes. The first one primary protocol PDC nondestructive quantum nondemolition (QND) detectors by transferring the spatial photon pairs to their polarization. In this time, QND act as role controlled-NOT (CNOT) gates. Also they can distinguish number modes, which provides good next process purify kept more. second purification, new are...
This paper aims at accelerating and compressing deep neural networks to deploy CNN models into small devices like mobile phones or embedded gadgets. We focus on filter level pruning, i.e., the whole will be discarded if it is less important. An effective unified framework, ThiNet (stands for "Thin Net"), proposed in this paper. formally establish pruning as an optimization problem, reveal that we need prune filters based statistics computed from its next layer, not current which...
Transformer, the model of choice for natural language processing, has drawn scant attention from medical imaging community. Given ability to exploit long-term dependencies, transformers are promising help atypical convolutional neural networks overcome their inherent shortcomings spatial inductive bias. However, most recently proposed transformer-based segmentation approaches simply treated as assisted modules encode global context into representations. To address this issue, we introduce...
Transformer, the model of choice for natural language processing, has drawn scant attention from medical imaging community. Given ability to exploit long-term dependencies, transformers are promising help atypical convolutional neural networks learn more contextualized visual representations. However, most recently proposed transformer-based segmentation approaches simply treated as assisted modules encode global context into To address this issue, we introduce nnFormer (i.e., not-another...
Graph Neural Networks (GNNs) have gained momentum in graph representation learning and boosted the state of art a variety areas, such as data mining (e.g., social network analysis recommender systems), computer vision object detection point cloud learning), natural language processing relation extraction sequence to name few. With emergence Transformers vision, embed structure into Transformer architecture overcome limitations local neighborhood aggregation while avoiding strict structural...
We present a nonlocal entanglement concentration scheme for reconstructing some maximally entangled multipartite states from partially ones by exploiting cross-Kerr nonlinearities to distinguish the parity of two polarization photons. Compared with schemes based on two-particle collective unitary evolution, this does not require parties know accurately information about states---i.e., their coefficients. Moreover, it possess sophisticated single-photon detectors, which makes protocol...
We propose a theoretical scheme for secure quantum key distribution network following the ideas in dense coding. In this scheme, server of provides service preparing and measuring Bell states, users encode states with local unitary operations. For preventing from eavesdropping, we design decoy when particle is transmitted between users. The has high capacity as one carries two bits information its efficiency qubits approaches 100%. Moreover, it unnecessary to store which makes more...
We present a scheme for symmetric multiparty quantum state sharing of an arbitrary m-qubit with m Greenberger–Horne–Zeilinger states following some ideas from the controlled teleportation (2005 Phys. Rev. A 72 02338). The sender Alice performs Bell-state measurements on her 2m particles and controllers need only take single-photon product their photons independently, not multipartite entanglement measurements, which makes this more convenient than latter. Also it does require parties to...
An efficient quantum cryptography network protocol is proposed with d-dimensional polarized photons, without resorting to entanglement and memory. A server on the network, say Alice, provides service for preparing measuring single photons whose initial state are |0⟩. The users code information some unitary operations. To prevent untrustworthy Alice from eavesdropping lines, a nonorthogonal-coding technique used in process that signal transmitted between users. This does not require servers...
This paper presents a scheme for quantum secure direct communication with encryption. The two authorized users use repeatedly sequence of the pure entangled pairs (quantum key) shared encrypting and decrypting secret message carried by travelling photons directly. For checking eavesdropping, parties perform single-photon measurements on some decoy particles before each round. has advantage that signal source is feasible at present any eavesdropper cannot steal message.