Audio processing by using pytorch 1D convolution network

kapre: Keras Audio Preprocessors

UrbanSound classification using Convolutional Recurrent Networks in PyTorch

:star: 本科毕业设计：基于内容的音乐推荐系统设计与开发。使用了Pytorch框架构建训练模型代码，使用Django构建了前后端。

Polish bird species recognition - Bird song analysis and classification with MFCC and CNNs. Trained on EfficientNets with final score 0.88 AUC. Women in Machine Learning & Data Science project.

From frequencies to feeling

C/C++实现Python音频处理库librosa中melspectrogram的计算过程

Perform three types of feature extraction: STFT, MFCC and MelSpectrogram. Apply CNN/VGG with or without RNN architecture. Able to achieve 95% accuracy.

A packaged convolutional voice activity detector for noisy environments.

Learnable STRF, from Riad et al. 2021 JASA

A C++ implementation of stft, melspectrogram and mel_to_stft

musical genres binary classification using pytorch.audio and keras

Signal Processing with Python and Librosa

Fashion Mnist and "recognize a speaker" datasets were utilized for image classification. For this classification task were tried to apply transfer learning from Mnist Fashion to "Recognize a Speaker" and transfer learning inside of Mnist Fashion.

A simple Speaker classifier using Keras

During the project for the DIGITAL SIGNAL IMAGE MANAGEMENT course I learned how to manage and process audio and image files. The aim of the project was the classification, through machine learning and deep learning models, of musical genres by extracting specific audio features from the "gtzan dataset" dataset files with which to train the models (SVM, Linear Regression, Decision tree , Random Forest, Neural Network). Mel spectograms were also extracted to train convolutional neural network models. In addition, the extracted audio features have been used to develop a model of music retrieval which given an audio track in input produces as output 5 audio tracks recommended meiante the use of cousine similarity.

Deep Learning project

An example repository to analyze cough audio data using transfer learning

audio classification fastai - Convert audio files into images for classification

This is a project on ML, Music Genre Classification using Mel-spectogram for audio preprocessing and Convolutional Neural Network(CNN) for the model training. Built CNN kinda scratch with no ml libraries but only using low-level tf operations.

The study presents a method to detect emotional states in real-time using audio data. It preprocesses the audio and trains a CNN to recognize emotions, achieving high accuracy. The system has potential applications in interactive systems and mental health monitoring, contributing to the development of emotionally intelligent technology.

In this project, I implemented Convolutional Neural Networks on images of melspectrogram of sound files.

This repository is mainly about the classification of music genres of Kaggle music dataset in various forms of data preparation

Sentiment Analysis from Speech!

CNN-LSTM model for audio emotion detection in children with adverse childhood events.

Classify audio recordings to a set of emotions.

Detecting emotions from audios using neural networks

Music recommendation using neural network

Trim or add an intro or outro to audio.

This repo explains and tests various machine learning models on the well-known GTZAN dataset and at the end compares them based on different metrics to choose the best one.