My photo

Siddharth Sigtia

Siri Speech, Apple
UK

sidsigtia at apple.com

Previously

PhD at EECS, QMUL
Supervisor: Prof. Simon Dixon

About me

I am a researcher on the Siri Speech team at Apple where I work on acoustic modelling with neural nets. Recently I have been working on the problem of keyword detection (e.g. detecting the phrase Hey Siri) in very low resource environments. More generally, I am interested in neural nets for acoustic and language modelling for automatic speech recognition and music signal processing.

Between 2012 and 2016, I was a PhD student at the Electronics Engineering and Computer Science department at Queen Mary where I was supervised by Prof. Simon Dixon. My PhD research was broadly in the field of Music Information Retrieval, where I investigated the use of neural network acoustic models for problems like automatic note and chord transcription. During the second half of my PhD, I also worked on the problem of Automatic Event Detection for non-speech, non-music sounds (or enviornmental sounds).

This webpage contains an up-to-date (hopefully) list of publications and any supplementary materials related to papers writted during my PhD. If you have any questions please feel free to email me at my work email address above.

Publications

Conferences

  • Efficient Voice Trigger Detection for Low Resource Hardware
     Siddharth Sigtia, Rob Haynes, Hywel Richards, Erik Marchi and John Bridle
    INTERSPEECH, Sep. 2018
    Paper
  • CHiME-home: A Dataset for Sound Source Recognition in a Domestic Environment
     Peter Foster, Siddharth Sigtia, Sacha Krstulovic, Jon Barker and Mark D Plumbley
    IEEE Workshop on Applications of Signal Processing to Audio and Acoustics (WASPAA), 2015
    Paper
  • A Hybrid Recurrent Neural Network for Music Transcription
     Siddharth Sigtia, Emmanouil Benetos, Nicolas Boulanger-Lewandowski, Tillman Weyde, Artur S d'Avila Garcez and Simon Dixon
    IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2015
    Paper,
  • Audio Chord Recognition with a Hybrid Recurrent Neural Network
     Siddharth Sigtia, Nicolas Boulanger-Lewandowski and Simon Dixon
    International Society for Music Information Retrieval (ISMIR), 2015
    Paper
  • Improved Music Feature Learning with Deep Neural Networks
     Siddharth Sigtia and Simon Dixon
    IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2015
    Paper
  • Learning to Generate Genotypes with Neural Networks
     Alexander W Churchill, Siddharth Sigtia and Chrisantha Fernando
    Arxiv Preprint
    Paper
    • A Denoising Autoencoder that Guides Stochastic Search
     Alexander W Churchill, Siddharth Sigtia and Chrisantha Fernando
    Arxiv Preprint
    Paper

Journals

  • Unsupervised Feature Learning Based on Deep Models for Environmental Audio Tagging
     Yong Xu, Qiang Huang, Wenwu Wang, Peter Foster, Siddharth Sigtia , Philip JB Jackson and Mark D Plumbley
    IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2017
    Paper
  • An End-to-End Neural Network for Polyphonic Piano Music Transcription
     Siddharth Sigtia, Emmanouil Benetos and Simon Dixon
    IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2016
    • Paper, Supplementary Materials
  • Automatic Environmental Sound Recognition: Performance Versus Computational Cost
     Siddharth Sigtia, Adam M Stark, Sacha Krstulović and Mark D Plumbley
    IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2016
    Paper

Thesis

  • Neural Networks for Analysing Music and Environmental Audio
     Sidhharth Sigtia
    Supervisor: Prof. Simon Dixon
    Thesis