MUSICAL SCALE IDENTIFICATION USING NEURAL NETWORKS -Lyndon Quadros.

Slides:



Advertisements
Similar presentations
Memristor in Learning Neural Networks
Advertisements

Face Recognition: A Convolutional Neural Network Approach
1 Image Classification MSc Image Processing Assignment March 2003.
Support Vector Machines
Music and Mathematics are they related?. What is Sound? Sound consists of vibrations of the air. In the air there are a large number of molecules moving.
Introduction The aim the project is to analyse non real time EEG (Electroencephalogram) signal using different mathematical models in Matlab to predict.
Major Scale Construction. Properties of a Major Scale Let’s look at a C Major scale What do you notice about this sequence of pitches?
Outcomes  Look at the theory of self-organisation.  Other self-organising networks  Look at examples of neural network applications.
RBF Neural Networks x x1 Examples inside circles 1 and 2 are of class +, examples outside both circles are of class – What NN does.
Classification of Music According to Genres Using Neural Networks, Genetic Algorithms and Fuzzy Systems.
Neural Networks Lab 5. What Is Neural Networks? Neural networks are composed of simple elements( Neurons) operating in parallel. Neural networks are composed.
Aula 4 Radial Basis Function Networks
ADAPTIVE ALGORITHMS IN VIBRATION DIAGNOSIS A.V. Tsurko Liahushevich S.I. − Associate Professor Belarusian State University of Informatics and Radioelectronics.
Spam? Not any more !! Detecting spam s using neural networks ECE/CS/ME 539 Project presentation Submitted by Sivanadyan, Thiagarajan.
Radial-Basis Function Networks
Face Recognition Using Neural Networks Presented By: Hadis Mohseni Leila Taghavi Atefeh Mirsafian.
Multiclass object recognition
Convolutional Neural Networks for Image Processing with Applications in Mobile Robotics By, Sruthi Moola.
West Virginia University
Presented by: Kamakhaya Argulewar Guided by: Prof. Shweta V. Jain
Audio classification Discriminating speech, music and environmental audio Rajas A. Sambhare ECE 539.
Soft Computing Lecture 20 Review of HIS Combined Numerical and Linguistic Knowledge Representation and Its Application to Medical Diagnosis.
December 5, 2012Introduction to Artificial Intelligence Lecture 20: Neural Network Application Design III 1 Example I: Predicting the Weather Since the.
Kumar Srijan ( ) Syed Ahsan( ). Problem Statement To create a Neural Networks based multiclass object classifier which can do rotation,
Intrusion Detection Using Hybrid Neural Networks Vishal Sevani ( )
Artificial Neural Nets and AI Connectionism Sub symbolic reasoning.
An Example of Course Project Face Identification.
Radial Basis Function Networks:
COMPARISON OF IMAGE ANALYSIS FOR THAI HANDWRITTEN CHARACTER RECOGNITION Olarik Surinta, chatklaw Jareanpon Department of Management Information System.
An Introduction to Support Vector Machine (SVM) Presenter : Ahey Date : 2007/07/20 The slides are based on lecture notes of Prof. 林智仁 and Daniel Yeung.
The Perceptron. Perceptron Pattern Classification One of the purposes that neural networks are used for is pattern classification. Once the neural network.
Handwritten Recognition with Neural Network Chatklaw Jareanpon, Olarik Surinta Mahasarakham University.
Spam Detection Ethan Grefe December 13, 2013.
Soft Computing Lecture 8 Using of perceptron for image recognition and forecasting.
Singer similarity / identification Francois Thibault MUMT 614B McGill University.
Music Genre Classification Alex Stabile. Example File
Protein motif extraction with neuro-fuzzy optimization Bill C. H. Chang and Author : Bill C. H. Chang and Saman K. Halgamuge Saman K. Halgamuge Adviser.
An Artificial Neural Network Approach to Surface Waviness Prediction in Surface Finishing Process by Chi Ngo ECE/ME 539 Class Project.
Chapter 3 Intervals and Transposition. Important Concepts Tone combinations are classified in music with names that identify the pitch relationships.
Neural Network Recognition of Frequency Disturbance Recorder Signals Stephen Tang REU Final Presentation July 22, 2014.
SUPPORT VECTOR MACHINES Presented by: Naman Fatehpuria Sumana Venkatesh.
Deep Belief Network Training Same greedy layer-wise approach First train lowest RBM (h 0 – h 1 ) using RBM update algorithm (note h 0 is x) Freeze weights.
Pattern Recognition. What is Pattern Recognition? Pattern recognition is a sub-topic of machine learning. PR is the science that concerns the description.
Feasibility of Using Machine Learning Algorithms to Determine Future Price Points of Stocks By: Alexander Dumont.
Classification of Breast Cancer Cells Using Artificial Neural Networks and Support Vector Machines Emmanuel Contreras Guzman.
Optical Character Recognition
A Document-Level Sentiment Analysis Approach Using Artificial Neural Network and Sentiment Lexicons Yan Zhu.
Neural network based hybrid computing model for wind speed prediction K. Gnana Sheela, S.N. Deepa Neurocomputing Volume 122, 25 December 2013, Pages 425–429.
Automatic Classification of Audio Data by Carlos H. L. Costa, Jaime D. Valle, Ro L. Koerich IEEE International Conference on Systems, Man, and Cybernetics.
Course Outline (6 Weeks) for Professor K.H Wong
Presentation on Artificial Neural Network Based Pathological Voice Classification Using MFCC Features Presenter: Subash Chandra Pakhrin 072MSI616 MSC in.
Other Classification Models: Neural Network
A Support Vector Machine Approach to Sonar Classification
Schizophrenia Classification Using
Self organizing networks
Dynamic Routing Using Inter Capsule Routing Protocol Between Capsules
Neural Networks: Improving Performance in X-ray Lithography Applications ECE 539 Ryan T. Hogg May 10, 2000.
Convolutional Neural Networks
Presented by Steven Lewis
شبکه عصبی تنظیم: بهروز نصرالهی-فریده امدادی استاد محترم: سرکار خانم کریمی دانشگاه آزاد اسلامی واحد شهرری.
Cache Replacement Scheme based on Back Propagation Neural Networks
network of simple neuron-like computing elements
An Improved Neural Network Algorithm for Classifying the Transmission Line Faults Slavko Vasilic Dr Mladen Kezunovic Texas A&M University.
Creating Data Representations
Department of Electrical Engineering
Example I: Predicting the Weather
Temporal Back-Propagation Algorithm
CSE 802. Prepared by Martin Law
Face Recognition: A Convolutional Neural Network Approach
Presentation transcript:

MUSICAL SCALE IDENTIFICATION USING NEURAL NETWORKS -Lyndon Quadros.

Scales and Octaves Musical scale - (music) a series of notes differing in pitch according to a specific scheme (usually within an octave) Octave-a series of eight notes occupying the interval between (and including) two notes, one having twice or half the frequency of vibration of the other.

Which Scales and Which Octaves? Western classical musical scales. All majors and their relative minors. A/F#m, A#/Gm, B/G#m, C/Am, D/Bm, D#/Cm, E/C#m, F/Dm, F#/D#m, G/Em, G#/Fm 12 Scales in all i.e. 12 Output Classes

Which Scales and Which Octaves? Scales in all the 10 octaves can be classified. However, Octaves -3 to +5 are audible to human ear. All the notes obtained will be scaled (normalised) down to the -3 octave.

Octave -3 Note NameFrequency Hz C C # /D b D D # /E b E F F # /G b G G # /A b A55 A # /B b B

Feature Vectors Frequencies of the notes that are present in each scale. Western music defines 7 notes for each scale. Hence, the input feature vector will be a 7- dimensional vector. Frequencies are obtained by Pitch detection

Pitch and Pitch Detection Pitch : The degree of “highness” or “lowness” of a note. Can be quantified in terms of frequency or number of cents from a reference note. Tarsos_Yin pitch detection algorithm has been employed

Training Inputs

Testing Data 36 vectors of 7 features (All 12 scales in three different progressions) Pre-processed to obtain the frequencies and extract the 7 most frequently occurring frequencies from the pitch detection. Normalised to the -3 frequency and arranged.

Current Status Completed feature extraction algorithm. Pattern classification using MLP and back propagation algorithm for the current set of data gives a maximum classification rate of 10.33% with 2 hidden layers of 14 neurons each and learning rate 0.1 Lesser classification rate due to lower dimensionality of input as compared to output.

Got a Job to be Done Since the lower dimensionality hinders classification, Radial Basis Networks and SVM appear to be the best options. Success is also subject to accurate pitch detection. Hence, various different pitch detection algorithms to be tested.

THANK YOU