Search

Your search keyword '"Li, Zhaoxia"' showing total 1,178 results
Start Over Author "Li, Zhaoxia"
1,178 results on '"Li, Zhaoxia"'

2. Multi-Task Sub-Band Network For Deep Residual Echo Suppression

Author

Sun, Jiayao, Luo, Dawei, Li, Zhaoxia, Li, Jindong, Ju, Yukai, and Li, Yang

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

This paper introduces the SWANT team entry to the ICASSP 2023 AEC Challenge. We submit a system that cascades a linear filter with a neural post-filter. Particularly, we adopt sub-band processing to handle full-band signals and shape the network with multi-task learning, where dual signal voice activity detection (DSVAD) and echo estimation are adopted as auxiliary tasks. Moreover, we particularly improve the time frequency convolution module (TFCM) to increase the receptive field using small convolution kernels. Finally, our system has ranked 4th in ICASSP 2023 AEC Challenge Non-personalized track.

Published
2023

6. The PCG-AIID System for L3DAS22 Challenge: MIMO and MISO convolutional recurrent Network for Multi Channel Speech Enhancement and Speech Recognition

Author

Li, Jingdong, Zhu, Yuanyuan, Luo, Dawei, Liu, Yun, Cui, Guohui, and Li, Zhaoxia

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing, Computer Science - Sound
Abstract

This paper described the PCG-AIID system for L3DAS22 challenge in Task 1: 3D speech enhancement in office reverberant environment. We proposed a two-stage framework to address multi-channel speech denoising and dereverberation. In the first stage, a multiple input and multiple output (MIMO) network is applied to remove background noise while maintaining the spatial characteristics of multi-channel signals. In the second stage, a multiple input and single output (MISO) network is applied to enhance the speech from desired direction and post-filtering. As a result, our system ranked 3rd place in ICASSP2022 L3DAS22 challenge and significantly outperforms the baseline system, while achieving 3.2% WER and 0.972 STOI on the blind test-set., Comment: To appear at ICASSP 2022 (Accepted)

Published
2022

13. FOXM1 promotes proliferation of lung adenocarcinoma cells by regulating HJURP

Author

ZHA Haoran and LI Zhaoxia

Subjects
hjurp, nsclc, foxm1, cancer cell proliferation, Medicine (General), R5-920
Abstract

Objective To investigate the relationship between the expression of Holliday junction recognition protein(HJURP) in lung cance tissues and prognosis of patients with lung adenocarcinoma(LUAD), and to explore the effect of HJURP on proliferation of LUAD cells(H1299 and SPC-A1). Methods Human non-small cell lung cancer (NSCLC) tissues and paired normal lung tissues(n=17) were collected from Department of Oncology of Rocket Force Medical Center from February 2019 to May 2020. RT-qPCR and Western blotting were applied to analyze the expression of HJURP in NSCLC tissues and normal tissues. Public available databases, the cancer genome atlas (TCGA) and Kaplan-Meier Plotter(KM-plot), were used to analyze the expression of HJURP of LUAD tissues and normal tissues and the predictive effect of prognosis in patients with LUAD. 5-ethynyl-2-deoxyuridine (EDU) incorporation assay was applied to investigate the effect of knock-dowm HJURP on the proliferation of cancer cell lines (H1299, SPC-A1). Pearson correlation analysis was used to analyze the relationship between HJURP and the expression of forkhead box protein M1(FOXM1). Chromatin immunoprecipitation(ChIP) assay was used to analyze the transcription regulation effect of FOXM1 on HJUPR. Dual-luciferase reporter assay was used to analyze the combination of FOXM1 and HJURP promoter. SiRNA was used to knock down FOXM1, which could detect the effect on the expression of HJURP. Results Compared with normal lung tissues, LUAD tissues had increased HJURP expression via RT-qPCR and Western blotting(P < 0.05). Increased HJURP expression in tumor tissues was correlated with a poor prognosis of patients with LUAD based on survival analysis of LUAD patients in TCGA and KM-plot(P < 0.05). Knock-down of HJURP inhibited the proliferation of H1299 (P < 0.05) and SPC-A1(P < 0.05) cells. The expression of FOXM1 correlated positively with HJURP in LUAD (medical center: r=0.581 3, P < 0.05; TCGA: r=0.877 6, P < 0.05; CCLE: r=0.455 1, P < 0.05). The results of ChIP assay and dual-luciferase reporter assay showed FOXM1 could combine the promoter region of HJURP. The expression of HJURP decreased after knocking down FOXM1(P < 0.05). Conclusion HJURP is involved in the regulation of the proliferation of LUAD cells. Increased HJURP expression in LUAD tissues is related to a poor prognosis. FOXM1 is the main upstream transcription factor for regulating HJURP.

Published
2023
Full Text
View/download PDF

37. Forecasting-aided state estimation of integrated energy systems based on improved particle filter

Author

YANG Dechang, WANG Yaning, LI Zhaoxia, GONG Xuejiao, YU Jianshu, and LI Ling

Subjects
integrated energy system, state estimation, particle filter algorithm, electricity-heat-gas network, tracking error, forecasting-aided, Applications of electric power, TK4001-4102
Abstract

Efficient and accurate state estimation is the basis for the safety and stability of the integrated energy system (IES). Particle filter has high precision and strong adaptability to nonlinear systems,and it has been applied to state estimation of power systems. To improve the precision of state estimation in IES,a forecasting-aided state estimation method based on improved particle filter is proposed. Firstly,a regional IES model including an electricity-heat-gas network is constructed. Secondly,the particle filter algorithm is applied to the electricity-heat-gas network. The prediction step of the particle filter is improved because of the tracking error problem of traditional particle filtering algorithm,which is based on particle filter theory. Finally,the improved particle filter algorithm is verified by using the classical IES example. The results show that this method can effectively solve the tracking error problem of the traditional particle filter algorithm,which can improve the precision of state estimation in IES.

Published
2022
Full Text
View/download PDF

41. Research on computer static software defect detection system based on big data technology

Author

Li Zhaoxia, Zhu Jianxing, Arumugam K., Bhola Jyoti, and Neware Rahul

Subjects
software defect prediction, detection system, big data technology, static software, software evolution, Science, Electronic computers. Computer science, QA75.5-76.95
Abstract

To study the static software defect detection system, based on the traditional static software defect detection system design, a new static software defect detection system design based on big data technology is proposed. The proposed method can optimize the distribution of test resources and improve the quality of software products by predicting the potential defect program modules and design the software and hardware of the static software defect detection system of big data technology. It is found that the traditional static software defect detection system design based on code source data takes a long time, averaging 65 h /day. However, the traditional static software defect detection system based on deep learning has a short detection time, averaging 35 h/day. In this article, the detection time of the static software defect detection system based on big data is shorter than that of the other two traditional system designs, with an average of 15 h/day. Because the system design adjusts the operating state of the system, it improves the accuracy of data operation. On the premise of data collection, the system inspection research is completed, which ensures the operational safety of software data, alleviates the contradiction between system and data to a high degree, improves the efficiency of system operation, reduces unnecessary operations, further shortens the time required for inspection, improves the system performance, and has higher research and operation value.

Published
2022
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results