Xiaoyan Fu

Bio: Xiaoyan Fu is an academic researcher. The author has contributed to research in topics: Computer science & Compressed sensing. The author has co-authored 3 publications.

Audio Signal-based Depression Level Prediction Combining Temporal and Spectral Features

Abstract: Depression is a serious mental disorder that affects millions of people worldwide. The prediction of depression level at early stage is significant. This paper focuses on predicting the degree of depression not just for judging depression based on audio signal. To induce dynamic temporal information of frequency domain, we proposed Audio Delta Ternary Patterns (ADTP) algorithm in the spectrogram feature space. Moreover, we designed an integrated model of multiple stream, which uses joint tuning layers to encode temporal movement feature, high-level features of spectral and MFCC, and predict the Beck Depression Inventory-II (BDI-II) values from speech signal. Experiments on the AVEC2014 dataset show that our method performs better than some previous methods in predicting depression scores based on audio data.

MRI Reconstruction using Minimax-Concave Total Variation Regularization based on p-norm

Abstract: Magnetic resonance imaging (MRI) reconstruction model based on total variation (TV) regularization can solve some problems, e.g., incomplete reconstruction, blurred imaging, and denoising. However, it has problems such as sensitivity to outliers, poor ability to induce the sparsity of the gradient domain of MR image. In this paper, minimax-concave total variation regularization based on $L_{p}-$norm (MCTV-Lp) is proposed to overcome these drawbacks. Specifically, the TV-Lp regularization is constructed using the exponent ${p}(0\lt{p}\lt 1)$, which is defined as the $L_{p}-$norm of the gradient. Then TV-Lp is combined with the minimax-concave penalty of the $L_{p}-$norm to construct the MCTV-Lp. Finally, the sparse reconstruction model based on minimax-concave total variation (MCTV-SRM) is proposed, where the objective function is formulated as the sum of the regularization of MCTV-Lp and the data-fitting term of $L_{2}-$norm. Moreover, an optimization algorithm based on the alternating direction method of multipliers (ADMM) is given to solve the related optimization problems iteratively. Results on different datasets with different experimental settings show that the proposed method is better adapted to MRI reconstruction and the relative error and PSNR are significantly improved than several typical methods, while can reconstruct MR images with clear details and textures.

Hankel Structured Low Rank and Sparse Representation Via L0-Norm Optimization for Compressed Ultrasound Plane Wave Signal Reconstruction

Abstract: Ultrasound plane wave imaging is widely used in many applications thanks to its capability in reaching high frame rates. However, the amount of data acquisition and storage in a period of time can become a bottleneck in ultrasound system design for thousands frames per second. In our previous study, we proposed a low-rank and joint-sparse model to reduce the amount of sampled channel data of focused beam imaging by considering all the received data as a 2D matrix. However, for a single plane wave transmission, the number of channels is limited and the low-rank property of the received data matrix is no longer achieved. In this study, a L 0 -norm based Hankel structured low-rank and sparse model is proposed to reduce the channel data. An optimization algorithm, based on the alternating direction method of multipliers (ADMM), is proposed to efficiently solve the resulting optimization problem. The performance of the proposed approach was evaluated using the data published in Plane Wave Imaging Challenge in Medical Ultrasound (PICMUS) in 2016. Results on channel and plane wave data show that the proposed method is better adapted to the ultrasound channel signal and can recover the image with fewer samples than the conventional CS method.