s603122001 / Vocal Melody Extraction
Licence: mit
Source code for "Vocal melody extraction with semantic segmentation and audio-symbolic domain transfer learning".
Stars: ✭ 44
Programming Languages
python
139335 projects - #7 most used programming language
Projects that are alternatives of or similar to Vocal Melody Extraction
Muspy
A toolkit for symbolic music generation
Stars: ✭ 151 (+243.18%)
Mutual labels: music-information-retrieval, music
Essentia
C++ library for audio and music analysis, description and synthesis, including Python bindings
Stars: ✭ 1,985 (+4411.36%)
Mutual labels: music-information-retrieval, music
Alignmentduration
Lyrics-to-audio-alignement system. Based on Machine Learning Algorithms: Hidden Markov Models with Viterbi forced alignment. The alignment is explicitly aware of durations of musical notes. The phonetic model are classified with MLP Deep Neural Network.
Stars: ✭ 36 (-18.18%)
Mutual labels: music-information-retrieval, music
Gist
A C++ Library for Audio Analysis
Stars: ✭ 244 (+454.55%)
Mutual labels: music-information-retrieval, music
Awesome Deep Learning Music
List of articles related to deep learning applied to music
Stars: ✭ 2,195 (+4888.64%)
Mutual labels: music-information-retrieval, music
Audioowl
Fast and simple music and audio analysis using RNN in Python 🕵️♀️ 🥁
Stars: ✭ 151 (+243.18%)
Mutual labels: music-information-retrieval, music
Music Synthesis With Python
Music Synthesis with Python talk, originally given at PyGotham 2017.
Stars: ✭ 48 (+9.09%)
Mutual labels: music-information-retrieval, music
Omr Datasets
Collection of datasets used for Optical Music Recognition
Stars: ✭ 158 (+259.09%)
Mutual labels: music-information-retrieval, music
Tutorial
Tutorial covering Open Source tools for Source Separation.
Stars: ✭ 223 (+406.82%)
Mutual labels: music-information-retrieval, music
Musicinformationretrieval.com
Instructional notebooks on music information retrieval.
Stars: ✭ 845 (+1820.45%)
Mutual labels: music-information-retrieval, music
Jeelangamusic
Discord bot with music functional. Play, skip, save music and etc!
Stars: ✭ 40 (-9.09%)
Mutual labels: music
Lofi Player
🎧 A Lofi Player built with HTML, CSS and Javascript using Parcel as Module Bundler https://lakscastro.github.io/lofi-player
Stars: ✭ 38 (-13.64%)
Mutual labels: music
Audioworks
A cross-platform, multi-format audio conversion and tagging suite
Stars: ✭ 35 (-20.45%)
Mutual labels: music
Musicode
🎶 Markup language for music creation and analysis! -- https://hlorenzi.github.io/musicode/
Stars: ✭ 34 (-22.73%)
Mutual labels: music
Ytspotifydl
Youtube and Spotify music downloader with metadata.
Stars: ✭ 34 (-22.73%)
Mutual labels: music
Starrysky
🔥A Powerful and Streamline MusicLibrary(一个丰富的音乐播放封装库,支持多种音频格式,完美解决你的问题。)
Stars: ✭ 1,022 (+2222.73%)
Mutual labels: music
Chainer Segnet
SegNet implementation & experiments in Chainer
Stars: ✭ 42 (-4.55%)
Mutual labels: semantic-segmentation
Vocal Melody Extraction
This repository includes the source code of the melody extraction algorithm from:
Wei-Tsung Lu and Li Su, “Vocal melody extraction with semantic segmentation and audio-symbolic domain transfer learning,” International Society of Music Information Retrieval Conference (ISMIR), September 2018.
Wei-Tsung Lu and Li Su, "Deep Learning Models for Melody Perception: An Investigation on Symbolic Music Data," Proc. Asia Pacific Signal and Infor. Proc. Asso. Annual Summit and Conf. (APSIPA ASC), November 2018.
Dependencies
This repository requires following packages:
- python 3.6
- numpy
- tensorflow
- keras
- mido
Usage
usage: VocalMelodyExtraction.py [-h][-p phase]
[-t model_type][-d data_type][-da dataset_path][-la label_path]
[-ms model_path_symbolic][-w window_width][-b batch_size_train][-e epoch]
[-n steps][-o output_model_name]
[-m model_path] [-i input_file][-bb batch_size_train]
required arguments:
-da dataset_path path to data set
-la label_path path to dataset label
-ms model_path_symbolic path to symbolic model
optional arguments:
-h
-p phase phase: training or testing (default: "testing)
-t model_type model type: seg or pnn (default: "seg")
-d data_type data type: audio or symbolic (default: "audio")
-w window_width width of the input feature (default: 128)
-b batch_size_train batch size during training (default: 12)
-e epoch number of epoch (default: 5)
-n steps number of step per epoch (default: 6000)
-o output_model_name name of the output model (default: "out")
-m model_path path to existing model (default: "Seg")
-i input_file path to input file (default: "train01.wav")
-bb batch_size_train batch size during testing (default: 10)
Pretrained Models
Click here to download the pretrained models.
Todos
- Add codes for symbolic model training
- Data set handling
License
MIT
Note that the project description data, including the texts, logos, images, and/or trademarks,
for each open source project belongs to its rightful owner.
If you wish to add or remove any projects, please contact us at [email protected].