Want to Generate your own Music using Deep Learning?

Here's a
Guide to do just that!
A D VA NC E D A UD I O A UD I O PRO C E S S I NG D E E P LE A RNI NG E NT E RT A I NM E NT PRO J E C T PYT HO N UNS T RUC T URE D D AT A

Overview

Learn how to develop an end-to-end model for Automatic Music Generation

Understand the WaveNet architecture and implement it from scratch using Keras
Compare the performance of WaveNet versus Long Short Term Memory for building an Automatic
Music Generation model

Introduction

“If I were not a physicist, I would probably be a musician. I often think in music. I live my daydreams in music. I see my life in terms of
music.” – Albert Einstein

I might not be a physicist like Mr. Einstein, but I wholeheartedly agree with his thoughts on music! I can’t
remember a single day when I didn’t open up my music player. My travel to and from office is accompanied
by the tune of music and honestly, it helps me focus on my work.

I’ve always dreamed of composing music but didn’t quite get the hang of instruments. That was until I
came across deep learning. Using certain techniques and frameworks, I was able to compose my own
original music score without really knowing any music theory!

This was one of my favorite professional projects. I combined my two passions – music and deep learning
– to create an automatic music generation model. It’s a dream come true!
I am thrilled to share my approach with you, including the entire code to enable you to generate your own
music! We’ll first quickly understand the concept of automatic music generation before diving into the
different approaches we can use to perform this. Finally, we will fire up Python and design our own
automatic music generation model.

Table of Contents

1. What is Automatic Music Generation?

2. What are the Constituent Elements of Music?
3. Different Approaches to Music Generation
1. Using WaveNet Architecture
2. Using Long Short Term Memory (LSTM)

4. Implementation – Automatic Music Composition using Python

What is Automatic Music Generation?

Music is an Art and a Universal language.

I define music as a collection of tones of different frequencies. So, the Automatic Music Generation is a
process of composing a short piece of music with minimum human intervention.
What could be the simplest form of generating music?

It all started by randomly selecting sounds and combining them to form a piece of music. In 1787, Mozart
proposed a Dice Game for these random sound selections. He composed nearly 272 tones manually! Then,
he selected a tone based on the sum of 2 dice.

Another interesting idea was to make use of musical grammar to generate music.

Musical Grammar comprehends the knowledge necessary to the just arrangement and combination of musical sounds and to the proper
performance of musical compositions
– Foundations of Musical Grammar

In the early 1950s, Iannis Xenakis used the concepts of Statistics and Probability to compose music –
popularly known as Stochastic Music. He defined music as a sequence of elements (or sounds) that
occurs by chance. Hence, he formulated it using stochastic theory. His random selection of elements was
strictly dependent on mathematical concepts.

Recently, Deep Learning architectures have become the state of the art for Automatic Music Generation. In
this article, I will discuss two different approaches for Automatic Music Composition using WaveNet and
LSTM (Long Short Term Memory) architectures.

Note: This article requires a basic understanding of a few deep learning concepts. I recommend going
through the below articles:

A Comprehensive Tutorial to learn Convolutional Neural Networks (CNNs) from Scratch

Essentials of Deep Learning: Introduction to Long Short Term Memory (LSTM)
Must-Read Tutorial to Learn Sequence Modeling

What are the Constituent Elements of Music?

Music is essentially composed of Notes and Chords. Let me explain these terms from the perspective of
the piano instrument:
 Note: The sound produced by a single key is called a note
Chords: The sound produced by 2 or more keys simultaneously is called a chord. Generally, most
chords contain at least 3 key sounds
Octave: A repeated pattern is called an octave. Each octave contains 7 white and 5 black keys

Different Approaches to Automatic Music Generation

I will discuss two Deep Learning-based architectures in detail for automatically generating music –
WaveNet and LSTM. But, why only Deep Learning architectures?

Deep Learning is a field of Machine Learning which is inspired by a neural structure. These networks
extract the features automatically from the dataset and are capable of learning any non-linear function.
That’s why Neural Networks are called as Universal Functional Approximators.

Hence, Deep Learning models are the state of the art in various fields like Natural Language Processing
(NLP), Computer Vision, Speech Synthesis and so on. Let’s see how we can build these models for music
composition.

Approach 1: Using WaveNet

WaveNet is a Deep Learning-based generative model for raw audio developed by Google DeepMind.

The main objective of WaveNet is to generate new samples from the original distribution of the data.
Hence, it is known as a Generative Model.

Wavenet is like a language model from NLP.

In a language model, given a sequence of words, the model tries to predict the next word. Similar to a
language model, in WaveNet, given a sequence of samples, it tries to predict the next sample.

Approach 2: Using Long Short Term Memory (LSTM) Model

Long Short Term Memory Model, popularly known as LSTM, is a variant of Recurrent Neural Networks
(RNNs) that is capable of capturing the long term dependencies in the input sequence. LSTM has a wide
range of applications in Sequence-to-Sequence modeling tasks like Speech Recognition, Text
Summarization, Video Classification, and so on.
Let’s discuss in detail how we can train our model using these two approaches.

Wavenet: The Training Phase

This is a Many-to-One problem where the input is a sequence of amplitude values and the output is the subsequent value.

Let’s see how we can prepare input and output sequences.

Input to the WaveNet:

WaveNet takes the chunk of a raw audio wave as an input. Raw audio wave refers to the representation of a
wave in the time series domain.

In the time-series domain, an audio wave is represented in the form of amplitude values which are recorded
at different intervals of time:

Output of the WaveNet:

Given the sequence of the amplitude values, WaveNet tries to predict the successive amplitude value.

Let’s understand this with the help of an example. Consider an audio wave of 5 seconds with a sampling
rate of 16,000 (that is 16,000 samples per second). Now, we have 80,000 samples recorded at different
intervals for 5 seconds. Let’s break the audio into chunks of equal size, say 1024 (which is a
hyperparameter).

The below diagram illustrates the input and output sequences for the model:
 Input and Output of first 3 chunks

We can follow a similar procedure for the rest of the chunks.

We can infer from the above that the output of every chunk depends only on the past information ( i.e.
previous timesteps) but not on the future timesteps. Hence, this task is known as Autoregressive task and
the model is known as an Autoregressive model.

Inference phase

In the inference phase, we will try to generate new samples. Let’s see how to do that:

1. Select a random array of sample values as a starting point to model

2. Now, the model outputs the probability distribution over all the samples
3. Choose the value with the maximum probability and append it to an array of samples
4. Delete the first element and pass as an input for the next iteration
5. Repeat steps 2 and 4 for a certain number of iterations

Understanding the WaveNet Architecture

The building blocks of WaveNet are Causal Dilated 1D Convolution layers. Let us first understand the
importance of the related concepts.

Why and What is a Convolution?

One of the main reasons for using convolution is to extract the features from an input.

For example, in the case of image processing, convolving the image with a filter gives us a feature map.
Convolution is a mathematical operation that combines 2 functions. In the case of image processing,
convolution is a linear combination of certain parts of an image with the kernel.

You can browse through the below article to read more about convolution:

 Architecture of Convolutional Neural Networks (CNNs) Demystified

What is 1D Convolution?

The objective of 1D convolution is similar to the Long Short Term Memory model. It is used to solve similar
tasks to those of LSTM. In 1D convolution, a kernel or a filter moves along only one direction:

The output of convolution depends upon the size of the kernel, input shape, type of padding, and stride.
Now, I will walk you through different types of padding for understanding the importance of using Dilated
Causal 1D Convolution layers.
When we set the padding valid, the input and output sequences vary in length. The length of an output is
less than an input:

When we set the padding to same, zeroes are padded on either side of the input sequence to make the
length of input and output equal:

Pros of 1D Convolution:

Captures the sequential information present in the input sequence

Training is much faster compared to GRU or LSTM because of the absence of recurrent connections

Cons of 1D Convolution:

When padding is set to the same, output at timestep t is convolved with the previous t-1 and future
timesteps t+1 too. Hence, it violates the Autoregressive principle
When padding is set to valid, input and output sequences vary in length which is required for
computing residual connections (which will be covered later)

This clears the way for the Causal Convolution.

Note: The pros and Cons I mentioned here are specific to this problem.

What is 1D Causal Convolution?

This is defined as convolutions where output at time t is convolved only with elements from time t and earlier in the previous layer.

In simpler terms, normal and causal convolutions differ only in padding. In causal convolution, zeroes are
added to the left of the input sequence to preserve the principle of autoregressive:
Pros of Causal 1D convolution:

Causal convolution does not take into account the future timesteps which is a criterion for building a
Generative model

Cons of Causal 1D convolution:

Causal convolution cannot look back into the past or the timesteps that occurred earlier in the
sequence. Hence, causal convolution has a very low receptive field. The receptive field of a network
refers to the number of inputs influencing an output:

As you can see here, the output is influenced by only 5 inputs. Hence, the Receptive field of the network is
5, which is very low. The receptive field of a network can also be increased by adding kernels of large sizes
but keep in mind that the computational complexity increases.

This drives us to the awesome concept of the Dilated 1D Causal Convolution.

What is Dilated 1D Causal Convolution?

A Causal 1D convolution layer with the holes or spaces in between the values of a kernel is known as Dilated 1D convolution.

The number of spaces to be added is given by the dilation rate. It defines the reception field of a network.
A kernel of size k and dilation rate d has d-1 holes in between every value in kernel k.

As you can see here, convolving a 3 * 3 kernel over a 7 * 7 input with dilation rate 2 has a reception field of
5 * 5.

Pros of Dilated 1D Causal Convolution:

 The dilated 1D convolution network increases the receptive field by exponentially increasing the
dilation rate at every hidden layer:

As you can see here, the output is influenced by all the inputs. Hence, the receptive field of the network is
16.

Residual Block of WaveNet:

A building block contains Residual and Skip connections which are just added to speed up the
convergence of the model:

The Workflow of WaveNet:

Input is fed into a causal 1D convolution

The output is then fed to 2 different dilated 1D convolution layers with sigmoid and tanh activations
The element-wise multiplication of 2 different activation values results in a skip connection
And the element-wise addition of a skip connection and output of causal 1D results in the residual

Long Short Term Memory (LSTM) Approach

Another approach for automatic music generation is based on the Long Short Term Memory (LSTM) model.
The preparation of input and output sequences is similar to WaveNet. At each timestep, an amplitude value
is fed into the Long Short Term Memory cell – it then computes the hidden vector and passes it on to the
next timesteps.

The current hidden vector at timestep h t is computed based on the current input a t and previously hidden
vector h t-1 . This is how the sequential information is captured in any Recurrent Neural Network:
Pros of LSTM:

Captures the sequential information present in the input sequence

Cons of LSTM:

It consumes a lot of time for training since it processes the inputs sequentially

Implementation – Automatic Music Generation using Python

The wait is over! Let’s develop an end-to-end model for the automatic generation of music. Fire up your
Jupyter notebooks or Colab (or whichever IDE you prefer).

Download the Dataset:

I downloaded and combined multiple classical music files of a digital piano from numerous resources. You
can download the final dataset from here.

Impor t libraries:

Music 21 is a Python library developed by MIT for understanding music data. MIDI is a standard format for
storing music files. MIDI stands for Musical Instrument Digital Interface. MIDI files contain the
instructions rather than the actual audio. Hence, it occupies very little memory. That’s why it is usually
preferred while transferring files.
#library for understanding music from music21 import *

Reading Musical Files:

Let’s define a function straight away for reading the MIDI files. It returns the array of notes and chords
present in the musical file.

1 #defining function to read MIDI files

2 def read_midi(file):
3
4 print("Loading Music File:",file)
5
6 notes=[]
7 notes_to_parse = None
8
9 #parsing a midi file
10 midi = converter.parse(file)
11
12 #grouping based on different instruments
13 s2 = instrument.partitionByInstrument(midi)
14
15 #Looping over all the instruments
16 for part in s2.parts:
 17
18 #select elements of only piano
19 if 'Piano' in str(part):
20
21 notes_to_parse = part.recurse()
22
23 #finding whether a particular element is note or a chord
24 for element in notes_to_parse:
25
26 #note
27 if isinstance(element, note.Note):
28 notes.append(str(element.pitch))
29
30 #chord
31 elif isinstance(element, chord.Chord):
32 notes.append('.'.join(str(n) for n in element.normalOrder))
33
34 return np.array(notes)

view raw
music_1.py hosted with ❤ by GitHub

Now, we will load the MIDI files into our environment

1 #for listing down the file names

2 import os
3
4 #Array Processing
5 import numpy as np
6
7 #specify the path
8 path='schubert/'
9
10 #read all the filenames
11 files=[i for i in os.listdir(path) if i.endswith(".mid")]
12
13 #reading each midi file
14 notes_array = np.array([read_midi(path+i) for i in files])

view raw
music_2.py hosted with ❤ by GitHub

Understanding the data:

Under this section, we will explore the dataset and understand it in detail.

1 #converting 2D array into 1D array

2 notes_ = [element for note_ in notes_array for element in note_]
3
4 #No. of unique notes
5 unique_notes = list(set(notes_))
6 print(len(unique_notes))

view raw
music_3.py hosted with ❤ by GitHub

Output: 304

As you can see here, no. of unique notes is 304. Now, let us see the distribution of the notes.

1 #importing library
2 from collections import Counter
3
4 #computing frequency of each note
5 freq = dict(Counter(notes_))
6
7 #library for visualiation
8 import matplotlib.pyplot as plt
9
10 #consider only the frequencies
11 no=[count for _,count in freq.items()]
12
 13 #set the figure size
14 plt.figure(figsize=(5,5))
15
16 #plot
17 plt.hist(no)

view raw
music_4.py hosted with ❤ by GitHub

Output:

From the above plot, we can infer that most of the notes have a very low frequency. So, let us keep the top
frequent notes and ignore the low-frequency ones. Here, I am defining the threshold as 50. Nevertheless,
the parameter can be changed.

frequent_notes = [note_ for note_, count in freq.items() if count>=50] print(len(frequent_notes))

Output: 167

As you can see here, no. of frequently occurring notes is around 170.  Now, let us prepare new musical
files which contain only the top frequent notes

1 new_music=[]
2
3 for notes in notes_array:
4 temp=[]
5 for note_ in notes:
6 if note_ in frequent_notes:
7 temp.append(note_)
8 new_music.append(temp)
9
10 new_music = np.array(new_music)

view raw
music_5.py hosted with ❤ by GitHub

Preparing Data:

Preparing the input and output sequences as mentioned in the article:

1 no_of_timesteps = 32
2 x = []
3 y = []
4
5 for note_ in new_music:
6 for i in range(0, len(note_) - no_of_timesteps, 1):
7
8 #preparing input and output sequences
9 input_ = note_[i:i + no_of_timesteps]
 10 output = note_[i + no_of_timesteps]
11
12 x.append(input_)
13 y.append(output)
14
15 x=np.array(x)
16 y=np.array(y)

view raw
music_6.py hosted with ❤ by GitHub

Now, we will assign a unique integer to every note:

unique_x = list(set(x.ravel())) x_note_to_int = dict((note_, number) for number, note_ in

enumerate(unique_x))

We will prepare the integer sequences for input data

1 #preparing input sequences

2 x_seq=[]
3 for i in x:
4 temp=[]
5 for j in i:
6 #assigning unique integer to every note
7 temp.append(x_note_to_int[j])
8 x_seq.append(temp)
9
10 x_seq = np.array(x_seq)

view raw
music_7.py hosted with ❤ by GitHub

Similarly, prepare the integer sequences for output data as well

unique_y = list(set(y)) y_note_to_int = dict((note_, number) for number, note_ in enumerate(unique_y))

y_seq=np.array([y_note_to_int[i] for i in y])

Let us preserve 80% of the data for training and the rest 20% for the evaluation:

from sklearn.model_selection import train_test_split x_tr, x_val, y_tr, y_val =

train_test_split(x_seq,y_seq,test_size=0.2,random_state=0)

Model Building

I have defined 2 architectures here – WaveNet and LSTM. Please experiment with both the architectures to
understand the importance of WaveNet architecture.

1 def lstm():
2 model = Sequential()
3 model.add(LSTM(128,return_sequences=True))
4 model.add(LSTM(128))
5 model.add(Dense(256))
6 model.add(Activation('relu'))
7 model.add(Dense(n_vocab))
8 model.add(Activation('softmax'))
9 model.compile(loss='sparse_categorical_crossentropy', optimizer='adam')
10 return model

view raw
lstm.py hosted with ❤ by GitHub

 
I have simplified the architecture of the WaveNet without adding residual and skip connections since the
role of these layers is to improve the faster convergence (and WaveNet takes raw audio wave as input). But
in our case, the input would be a set of nodes and chords since we are generating music:

1 from keras.layers import *

2 from keras.models import *
3 from keras.callbacks import *
4 import keras.backend as K
5
6 K.clear_session()
7 model = Sequential()
8
9 #embedding layer
10 model.add(Embedding(len(unique_x), 100, input_length=32,trainable=True))
11
12 model.add(Conv1D(64,3, padding='causal',activation='relu'))
13 model.add(Dropout(0.2))
14 model.add(MaxPool1D(2))
15
16 model.add(Conv1D(128,3,activation='relu',dilation_rate=2,padding='causal'))
17 model.add(Dropout(0.2))
18 model.add(MaxPool1D(2))
19
20 model.add(Conv1D(256,3,activation='relu',dilation_rate=4,padding='causal'))
21 model.add(Dropout(0.2))
22 model.add(MaxPool1D(2))
23
24 #model.add(Conv1D(256,5,activation='relu'))
25 model.add(GlobalMaxPool1D())
26
27 model.add(Dense(256, activation='relu'))
28 model.add(Dense(len(unique_y), activation='softmax'))
29
30 model.compile(loss='sparse_categorical_crossentropy', optimizer='adam')
31
32 model.summary()

view raw
10_8.py hosted with ❤ by GitHub

Define the callback to save the best model during training:

mc=ModelCheckpoint('best_model.h5', monitor='val_loss', mode='min', save_best_only=True,verbose=1)

Let’s train the model with a batch size of 128 for 50 epochs:

history = model.fit(np.array(x_tr),np.array(y_tr),batch_size=128,epochs=50, validation_data=

(np.array(x_val),np.array(y_val)),verbose=1, callbacks=[mc])

Loading the best model:

#loading best model from keras.models import load_model model = load_model('best_model.h5')

Its time to compose our own music now. We will follow the steps mentioned under the inference phase for
the predictions.

1 import random
2 ind = np.random.randint(0,len(x_val)-1)
3
4 random_music = x_val[ind]
5
6 predictions=[]
7 for i in range(10):
8
9 random_music = random_music.reshape(1,no_of_timesteps)
10
 11 prob = model.predict(random_music)[0]
12 y_pred= np.argmax(prob,axis=0)
13 predictions.append(y_pred)
14
15 random_music = np.insert(random_music[0],len(random_music[0]),y_pred)
16 random_music = random_music[1:]
17
18 print(predictions)

view raw
10_9.py hosted with ❤ by GitHub

Now, we will convert the integers back into the notes.

x_int_to_note = dict((number, note_) for number, note_ in enumerate(unique_x)) predicted_notes =

[x_int_to_note[i] for i in predictions]

The final step is to convert back the predictions into a MIDI file. Let’s define the function to accomplish the
task.

1 def convert_to_midi(prediction_output):
2
3 offset = 0
4 output_notes = []
5
6 # create note and chord objects based on the values generated by the model
7 for pattern in prediction_output:
8
9 # pattern is a chord
10 if ('.' in pattern) or pattern.isdigit():
11 notes_in_chord = pattern.split('.')
12 notes = []
13 for current_note in notes_in_chord:
14
15 cn=int(current_note)
16 new_note = note.Note(cn)
17 new_note.storedInstrument = instrument.Piano()
18 notes.append(new_note)
19
20 new_chord = chord.Chord(notes)
21 new_chord.offset = offset
22 output_notes.append(new_chord)
23
24 # pattern is a note
25 else:
26
27 new_note = note.Note(pattern)
28 new_note.offset = offset
29 new_note.storedInstrument = instrument.Piano()
30 output_notes.append(new_note)
31
32 # increase offset each iteration so that notes do not stack
33 offset += 1
34 midi_stream = stream.Stream(output_notes)
35 midi_stream.write('midi', fp='music.mid')

view raw
10_10.py hosted with ❤ by GitHub

Converting the predictions into a musical file:

convert_to_midi(predicted_notes)

Some of the tunes composed by the model:

0:00 / 0:00

0:00 / 0:00
 0:00 / 0:00

Awesome, right? But your learning doesn’t stop here. Just remember that we have built a baseline model.
There are plenty of ways to improve the performance of the model even further:

As the size of the training dataset is small, we can fine-tune a pre-trained model to build a robust
system
Collect as much as training data as you can since the deep learning model generalizes well on the
larger datasets

End Notes

Deep Learning has a wide range of applications in our daily life. The key steps in solving any problem are
understanding the problem statement, formulating it and defining the architecture to solve the problem.

I had a lot of fun (and learning) while working on this project. Music is a passion of mine and it was quite
intriguing combining deep learning with that.

I am looking forward to hearing your approach to the problem in the comments section. And if you have
any feedback on this article or any doubts/queries, kindly share them in the comments section below and I
will get back to you.

Article Url - https://www.analyticsvidhya.com/blog/2020/01/how-to-perform-automatic-music-generation/

Aravind Pai
Aravind is a sports fanatic. His passion lies in developing data-driven products for the sports domain.
He strongly believes that analytics in sports can be a game-changer