yoyoyo's starred repositories
Frequency_ridge_tracking
Frequency tracking in time-frequency representations
ssqueezepy
Synchrosqueezing, wavelet transforms, and time-frequency analysis in Python
DL-based-Intelligent-Diagnosis-Benchmark
Source codes for the paper "Deep Learning Algorithms for Rotating Machinery Intelligent Diagnosis: An Open Source Benchmark Study"
Signal-Processing
Processing Of A Seismic Signal Using Fourier Hilbert And Hilbert-Huang Transform in Python
Seismic_Sensory_Data_Analysis
Seismic data reconstruction is an important research direction in the field of seismic signal analysis. The complete seismic data can be used to estimate interior images of the Earth, which can aid the exploration for resources and research in to the shallow structure of the crust for geological and environmental purposes. However, due to the severely corrupted seismic traces and seismic slices, harsh detection conditions, and even financial constraints, seismic data usually has lots of missing data entries and noise. Therefore, it is necessary to investigate the robust recovery of seismic data from incomplete and noisy data.
Unsupervised_Deep_Learning
Unsupervised (Self-Supervised) Clustering of Seismic Signals Using Deep Convolutional Autoencoders
Speech-separation
A method of using time frequency analysis and deep learning
Underwater-Acoustic-Target-Classification-Based-on-Dense-Convolutional-Neural-Network
In oceanic remote sensing operations, underwater acoustic target recognition is always a difficult and extremely important task of sonar systems, especially in the condition of complex sound wave propagation characteristics. Expensively learning recognition model for big data analysis is typically an obstacle for most traditional machine learning (ML) algorithms, whereas convolutional neural network (CNN), a type of deep neural network, can automatically extract features for accurate classification. In this study, we propose an approach using a dense CNN model for underwater target recognition. The network architecture is designed to cleverly re-use all former feature maps to optimize classification rate under various impaired conditions while satisfying low computational cost. In addition, instead of using time-frequency spectrogram images, the proposed scheme allows directly utilizing original audio signal in time domain as the network input data. Based on the experimental results evaluated on the real-world dataset of passive sonar, our classification model achieves the overall accuracy of 98.85$\%$ at 0 dB signal-to-noise ratio (SNR) and outperforms traditional ML techniques, as well as other state-of-the-art CNN models.
TFC-pretraining
Self-supervised contrastive learning for time series via time-frequency consistency
DTL_TFC_Vibration_Identification
Deep Transfer Learning and Time-Frequency Characteristics-Based Identification Method for Structural Seismic Response