Beijing University of Posts and Telecommunications

About: Beijing University of Posts and Telecommunications is a education organization based out in Beijing, Beijing, China. It is known for research contribution in the topics: MIMO & Quality of service. The organization has 39576 authors who have published 41525 publications receiving 403759 citations. The organization is also known as: BUPT.

Submicron bidirectional all-optical plasmonic switches

Abstract: Ultra-small all-optical switches are of importance in highly integrated optical communication and computing networks. However, the weak nonlinear light-matter interactions in natural materials present an enormous challenge to realize efficiently switching for the ultra-short interaction lengths. Here, we experimentally demonstrate a submicron bidirectional all-optical plasmonic switch with an asymmetric T-shape single slit. Sharp asymmetric spectra as well as significant field enhancements (about 18 times that in the conventional slit case) occur in the symmetry-breaking structure. Consequently, both of the surface plasmon polaritons propagating in the opposite directions on the metal surface are all-optically controlled inversely at the same time with the on/off switching ratios of >6 dB for the device lateral dimension of <1 μm. Moreover, in such a submicron structure, the coupling of free-space light and the on-chip bidirectional switching are integrated together. This submicron bidirectional all-optical switch may find important applications in the highly integrated plasmonic circuits.

Breast cancer histopathology image classification through assembling multiple compact CNNs

Abstract: Breast cancer causes hundreds of thousands of deaths each year worldwide. The early stage diagnosis and treatment can significantly reduce the mortality rate. However, the traditional manual diagnosis needs intense workload, and diagnostic errors are prone to happen with the prolonged work of pathologists. Automatic histopathology image recognition plays a key role in speeding up diagnosis and improving the quality of diagnosis. In this work, we propose a breast cancer histopathology image classification by assembling multiple compact Convolutional Neural Networks (CNNs). First, a hybrid CNN architecture is designed, which contains a global model branch and a local model branch. By local voting and two-branch information merging, our hybrid model obtains stronger representation ability. Second, by embedding the proposed Squeeze-Excitation-Pruning (SEP) block into our hybrid model, the channel importance can be learned and the redundant channels are thus removed. The proposed channel pruning scheme can decrease the risk of overfitting and produce higher accuracy with the same model size. At last, with different data partition and composition, we build multiple models and assemble them together to further enhance the model generalization ability. Experimental results show that in public BreaKHis dataset, our proposed hybrid model achieves comparable performance with the state-of-the-art. By adopting the multi-model assembling scheme, our method outperforms the state-of-the-art in both patient level and image level accuracy for BACH dataset. We propose a novel compact breast cancer histopathology image classification scheme by assembling multiple compact hybrid CNNs. The proposed scheme achieves promising results for the breast cancer image classification task. Our method can be used in breast cancer auxiliary diagnostic scenario, and it can reduce the workload of pathologists as well as improve the quality of diagnosis.

Mitigating Bias in Face Recognition Using Skewness-Aware Reinforcement Learning

Abstract: Racial equality is an important theme of international human rights law, but it has been largely obscured when the overall face recognition accuracy is pursued blindly. More facts indicate racial bias indeed degrades the fairness of recognition system and the error rates on non-Caucasians are usually much higher than Caucasians. To encourage fairness, we introduce the idea of adaptive margin to learn balanced performance for different races based on large margin losses. A reinforcement learning based race balance network (RL-RBN) is proposed. We formulate the process of finding the optimal margins for non-Caucasians as a Markov decision process and employ deep Q-learning to learn policies for an agent to select appropriate margin by approximating the Q-value function. Guided by the agent, the skewness of feature scatter between races can be reduced. Besides, we provide two ethnicity aware training datasets, called BUPT-Globalface and BUPT-Balancedface dataset, which can be utilized to study racial bias from both data and algorithm aspects. Extensive experiments on RFW database show that RL-RBN successfully mitigates racial bias and learns more balanced performance.

Exact Outage Performance of Cognitive Relay Networks with Maximum Transmit Power Limits

Abstract: In this letter, considering a cognitive relay network with the maximum transmit power limit in a spectrum sharing scenario, the exact outage probability of secondary system is derived over Rayleigh fading channels, and the theoretical analysis is validated by simulations. The results show that both the maximum transmit power limit and the interference power constraint cause the outage saturation phenomenon, and more relays for secondary system can provide better outage performance.

Quantum private query: A new kind of practical quantum cryptographic protocol

Abstract: This research aims to review the developments in the field of quantum private query (QPQ), a type of practical quantum cryptographic protocol. The primary protocol, as proposed by Jacobi et al., and the improvements in the protocol are introduced. Then, the advancements made in sability, theoretical security, and practical security are summarized. Additionally, we describe two new results concerning QPQ security. We emphasize that a procedure to detect outside adversaries is necessary for QPQ, as well as for other quantum secure computation protocols, and then briefly propose such a strategy. Furthermore, we show that the shift-and-addition or low-shift-and-addition technique can be used to obtain a secure real-world implementation of QPQ, where a weak coherent source is used instead of an ideal single-photon source.