MIDI-Images / midi_images_solo_piano_dataset_maker.py
asigalov61's picture
Upload 4 files
e2fb2a2 verified
raw
history blame
8.56 kB
# -*- coding: utf-8 -*-
"""MIDI_Images_Solo_Piano_Dataset_Maker.ipynb
Automatically generated by Colab.
Original file is located at
https://colab.research.google.com/drive/15E6o3Y1xPific5RtIZ-1CneHQhts3eEr
# MIDI Images Solo Piano Dataset Maker (ver. 1.0)
***
Powered by tegridy-tools: https://github.com/asigalov61/tegridy-tools
***
#### Project Los Angeles
#### Tegridy Code 2024
***
# (SETUP ENVIRONMENT)
"""
# @title Install dependecies
!git clone --depth 1 https://github.com/asigalov61/tegridy-tools
# Commented out IPython magic to ensure Python compatibility.
#@title Import all needed modules
print('=' * 70)
print('Loading core modules...')
print('Please wait...')
print('=' * 70)
import os
import copy
import math
import statistics
import random
import pickle
import shutil
from itertools import groupby
from collections import Counter
from sklearn.metrics import pairwise_distances
from sklearn import metrics
from joblib import Parallel, delayed, parallel_config
import numpy as np
from tqdm import tqdm
from PIL import Image
import matplotlib.pyplot as plt
print('Done!')
print('=' * 70)
print('Creating I/O dirs...')
if not os.path.exists('/content/Dataset'):
os.makedirs('/content/Dataset')
print('Done!')
print('=' * 70)
print('Loading tegridy-tools modules...')
print('=' * 70)
# %cd /content/tegridy-tools/tegridy-tools
import TMIDIX
import TMELODIES
import TPLOTS
import HaystackSearch
# %cd /content/
print('=' * 70)
print('Done!')
print('=' * 70)
"""# (DOWNLOAD SAMPLE MIDI DATASET)"""
# Commented out IPython magic to ensure Python compatibility.
# @title Download sample MIDI dataset (POP909)
# %cd /content/Dataset/
!git clone --depth 1 https://github.com/music-x-lab/POP909-Dataset
# %cd /content/
#@title Save file list
###########
print('=' * 70)
print('Loading MIDI files...')
print('This may take a while on a large dataset in particular...')
dataset_addr = '/content/Dataset/'
# os.chdir(dataset_addr)
filez = list()
for (dirpath, dirnames, filenames) in os.walk(dataset_addr):
filez += [os.path.join(dirpath, file) for file in filenames if file.endswith('.mid') or file.endswith('.midi') or file.endswith('.kar')]
print('=' * 70)
if filez == []:
print('Could not find any MIDI files. Please check Dataset dir...')
print('=' * 70)
print('Randomizing file list...')
random.shuffle(filez)
print('Done!')
print('=' * 70)
print('Total found MIDI files:', len(filez))
print('=' * 70)
TMIDIX.Tegridy_Any_Pickle_File_Writer(filez, 'filez')
print('=' * 70)
"""# (LOAD TMIDIX MIDI PROCESSOR)"""
#@title Load TMIDIX MIDI processor
print('=' * 70)
print('TMIDIX MIDI Processor')
print('=' * 70)
print('Loading...')
###########
def TMIDIX_MIDI_Processor(midi_file):
fn = os.path.basename(midi_file)
fn1 = fn.split('.mid')[0]
try:
#=======================================================
# START PROCESSING
raw_score = TMIDIX.midi2single_track_ms_score(midi_file)
escore_notes = TMIDIX.advanced_score_processor(raw_score, return_enhanced_score_notes=True)[0]
escore_notes = TMIDIX.augment_enhanced_score_notes(escore_notes, timings_divider=256)
sp_escore_notes = TMIDIX.recalculate_score_timings(TMIDIX.solo_piano_escore_notes(escore_notes, keep_drums=False))
if sp_escore_notes:
bmatrix = TMIDIX.escore_notes_to_binary_matrix(sp_escore_notes)
return [fn1, bmatrix]
else:
return [fn1, []]
#=======================================================
except Exception as ex:
print('WARNING !!!')
print('=' * 70)
print('Bad MIDI:', midi_file)
print('Error detected:', ex)
print('=' * 70)
return None
print('Done!')
print('=' * 70)
"""# (PROCESS MIDIs)"""
#@title Process MIDIs with TMIDIX MIDI processor
output_folder = "/content/MIDI-Images/" # @param {"type":"string"}
NUMBER_OF_PARALLEL_JOBS = 4 # Number of parallel jobs
NUMBER_OF_FILES_PER_ITERATION = 4 # Number of files to queue for each parallel iteration
SAVE_EVERY_NUMBER_OF_ITERATIONS = 128 # Save every 2560 files
print('=' * 70)
print('TMIDIX MIDI Processor')
print('=' * 70)
print('Starting up...')
print('=' * 70)
###########
melody_chords_f = []
files_count = 0
print('Processing MIDI files...')
print('Please wait...')
print('=' * 70)
for i in tqdm(range(0, len(filez), NUMBER_OF_FILES_PER_ITERATION)):
with parallel_config(backend='threading', n_jobs=NUMBER_OF_PARALLEL_JOBS, verbose = 0):
output = Parallel(n_jobs=NUMBER_OF_PARALLEL_JOBS, verbose=0)(delayed(TMIDIX_MIDI_Processor)(f) for f in filez[i:i+NUMBER_OF_FILES_PER_ITERATION])
for o in output:
if o is not None:
melody_chords_f.append(o)
if i % (NUMBER_OF_FILES_PER_ITERATION * SAVE_EVERY_NUMBER_OF_ITERATIONS) == 0 and i != 0:
print('SAVING !!!')
print('=' * 70)
print('Saving processed files...')
files_count += len(melody_chords_f)
print('=' * 70)
print('Processed so far:', files_count, 'out of', len(filez), '===', files_count / len(filez), 'good files ratio')
print('=' * 70)
print('Writing images...')
print('Please wait...')
for mat in melody_chords_f:
if mat[1]:
TPLOTS.binary_matrix_to_images(mat[1],
128,
32,
output_folder=output_folder+str(mat[0])+'/',
output_img_prefix=str(mat[0]),
output_img_ext='.png',
verbose=False
)
print('Done!')
print('=' * 70)
melody_chords_f = []
print('SAVING !!!')
print('=' * 70)
print('Saving processed files...')
files_count += len(melody_chords_f)
print('=' * 70)
print('Processed so far:', files_count, 'out of', len(filez), '===', files_count / len(filez), 'good files ratio')
print('=' * 70)
print('Writing images...')
print('Please wait...')
for mat in melody_chords_f:
if mat[1]:
TPLOTS.binary_matrix_to_images(mat[1],
128,
32,
output_folder=output_folder+str(mat[0])+'/',
output_img_prefix=str(mat[0]),
output_img_ext='.png',
verbose=False
)
print('Done!')
print('=' * 70)
"""# (LOAD IMAGES)"""
#@title Load created MIDI images
full_path_to_metadata_pickle_files = "/content/MIDI-Images" #@param {type:"string"}
print('=' * 70)
print('MIDI Images Reader')
print('=' * 70)
print('Searching for images...')
filez = list()
for (dirpath, dirnames, filenames) in os.walk(full_path_to_metadata_pickle_files):
filez += [os.path.join(dirpath, file) for file in filenames if file.endswith('.png')]
print('=' * 70)
filez.sort()
print('Found', len(filez), 'images!')
print('=' * 70)
print('Reading images...')
print('Please wait...')
print('=' * 70)
fidx = 0
all_read_images = []
for img in tqdm(filez):
img = Image.open(img)
img_arr = np.array(img).tolist()
all_read_images.append(img_arr)
fidx += 1
print('Done!')
print('=' * 70)
print('Loaded', fidx, 'images!')
print('=' * 70)
print('Done!')
print('=' * 70)
"""# (TEST IMAGES)"""
# @title Test created MIDI images
print('=' * 70)
image = random.choice(all_read_images)
escore = TMIDIX.binary_matrix_to_original_escore_notes(image)
output_score, patches, overflow_patches = TMIDIX.patch_enhanced_score_notes(escore)
detailed_stats = TMIDIX.Tegridy_ms_SONG_to_MIDI_Converter(output_score,
output_signature = 'MIDI Images',
output_file_name = '/content/MIDI-Images-Composition',
track_name='Project Los Angeles',
list_of_MIDI_patches=patches,
timings_multiplier=256
)
print('=' * 70)
"""# (ZIP IMAGES)"""
# @title Zip created MIDI images
!zip -9 -r POP909_MIDI_Images_128_128_32_BW.zip MIDI-Images/ > /dev/null
"""# Congrats! You did it! :)"""