I am a PhD candidate in the Signal Processing and Analysis Group at the Ming Hsieh Department of Electrical Engineering, USC. My research interests are in developing robust speech processing and analysis techniques for studying child-adult dyadic interactions. My supervisor is Prof. Shrikanth Narayanan.

Before joining SAIL, I received a Dual Degree (BTech + MTech) from IIT Madras, Chennai. I worked in the CompMusic project as part of my MTech project where I developed group delay based onset detection techniques for music (percussion) instruments. My project supervisor was Prof. Hema Murthy and my thesis is available here.

Code

• PyTorch Xvectors Deep speaker embeddings from audio

• ASR Models for child speech

• A python toolkit for speaker diarization based on Information bottleneck criterion

Preprints

• Designing Neural Speaker Embeddings with Meta Learning
  Manoj Kumar, Tae-Jin Park, Somer Bishop, Shrikanth Narayanan
  arXiv preprint arXiv:2207.16196(2020)
  


• Meta-learning with Latent Space Clustering in Generative Adversarial Network for Speaker Diarization
  Monisankha Pal, Manoj Kumar, Raghuveer Peri, Tae-Jin Park, So-Hyun Kim, Catherine Lord, Somer Bishop, Shrikanth Narayanan
  arXiv preprint arXiv:2007.09635(2020)
  


• A Study of Semi-supervised Speaker Diarization System using GAN Mixture Model
  Monisankha Pal, Manoj Kumar, Raghuveer Peri, Shrikanth Narayanan
  arXiv preprint arXiv:1910.11416(2019)
  


• Measuring Conversational Productivity in Child Forensic Interviews
  Victor Ardulov, Manoj Kumar, Shanna Williams, Thomas Lyon, Shrikanth Narayanan
  arXiv preprint arXiv:1806.03357 (2018)
  

Journal Publications

• Improving Speaker Diarization for Naturalistic Child-Adult Conversational Interactions using Contextual Information
  Manoj Kumar, So Hyun Kim, Catherine Lord, Shrikanth Narayanan
  Journal of Acoustical Society of America (2020)
  


• Leveraging Linguistic Context in Dyadic Interactions to Improve Automatic Speech Recognition for Children
  Manoj Kumar So-Hyun Kim, Catherine Lord, Thomas Lyon, Shrikanth Narayanan
  Computer Speech & Language (2020)
  


• Auto-tuning spectral clustering for speaker diarization using normalized maximum eigengap
  Tae-jin Park, Kyu Han, Manoj Kumar, Shrikanth Narayanan
  IEEE Signal Processing Letters (2019)
  


• An analysis of the high resolution property of group delay function with applications to audio signal processing
  Jilt Sebastian, Manoj Kumar, Hema A. Murthy
  Speech Communication (2016)
  [Link]

Conference Proceedings

• Meta-learning for Robust Child-Adult Classification from Speech
  Nithin Rao Koluguri, Manoj Kumar, So Hyun Kim, Catherine Lord, Shrikanth Narayanan
  ICASSP (2020)
  


• Learning Domain Invariant Representations for Child-Adult Classification from Speech
  Rimita Lahiri, Manoj Kumar, Somer Bishop, Shrikanth Narayanan
  ICASSP (2020)
  


• Speaker Clustering using Latent Space Clustering in Generative Adversarial Network
  Monisankha Pal, Manoj Kumar, Raghuveer Peri, Tae Jin Park, So Hyun Kim, Catherine Lord, Somer Bishop, Shrikanth Narayanan
  ICASSP (2020)
  


• Prototypical Networks for Robust Automatic Child-Adult Classification from Speech
  Manoj Kumar, Nithin Koluguri, So Hyun Kim, Catherine Lord and Shrikanth Narayanan
  International Society for Autism Research (INSAR) Annual Meeting (2020)
  


• The Second DIHARD challenge: System Description for USC-SAIL Team
  Taejin Park, Manoj Kumar, Nikolaos Flemotomos, Monisankha Pal, Raghuveer Peri, Rimita Lahiri, Panayiotis Georgiou and Shrikanth Narayanan
  Interspeech (2019)
  


• Robustness Analysis for Computational Speech Features during Naturalistic Clinician-Child Interactions
  Manoj Kumar, Karan Singla, Gabrielle Gunn, Catherine Lord, So Hyun Kim and Shrikanth Narayanan
  International Society for Autism Research (INSAR) Annual Meeting (2019)
  


• Multimodal Interaction Modeling of Child Forensic Interviewing
  Victor Ardulov, Madelyn Mendlen, Manoj Kumar, Neha Anand, Shanna Williams, Thomas Lyon, Shrikanth Narayanan
  ICMI (2018)
  


• Denoising and Raw-waveform Networks for Weakly-Supervised Gender Identification on Noisy Speech
  Jilt Sebastian, Manoj Kumar, D. S. Pavan Kumar, Mathew Magimai-Doss, Hema A. Murthy, Shrikanth Narayanan
  Interspeech (2018)
  


• A Knowledge Driven Structural Segmentation Approach for Play-Talk Classification During Autism Assessment
  Manoj Kumar, Pooja Chebolu, So Hyun Kim, Kassandra Martinez, Catherine Lord, Shrikanth Narayanan
  Interspeech (2018)
  


• Improving Semi-supervised Classification for Low-Resource Speech Interaction Applications
  Manoj Kumar, Pavlos Papadopoulos, Ruchir Travadi, Daniel Bone, Shrikanth Narayanan
  ICASSP (2018)
  [Poster], [Full Paper]


• Multi-scale Context Adaptation for Improving Child Automatic Speech Recognition in Child-Adult Spoken Interactions
  Manoj Kumar, Daniel Bone, Kelly McWilliams, Shanna Williams, Thomas Lyon, Shrikanth Narayanan
  Interspeech (2017)
  [Slides], [Full Paper]


• Objective Language Feature Analysis in Children with Neurodevelopmental Disorders during Autism Assessment
  Manoj Kumar, Rahul Gupta, Daniel Bone, Nikolaos Malandrakis, Somer Bishop, Shrikanth Narayanan
  Interspeech (2016)
  [Poster], [Full Paper]


• Musical Onset Detection on Carnatic Percussion Instruments
  Manoj Kumar, Jilt Sebastian, Hema A. Murthy
  NCC (2015)
  [Link]


• Pitch Estimation From Speech Using Grating Compression Transform on Modified Group-Delay-gram
  Jilt Sebastian, Manoj Kumar, Hema A. Murthy
  NCC (2015)
  [Link]


• Discovery of Syllabic Percussion Patterns in Tabla Solo Recordings
  Swapnil Gupta, Ajay Srinivasamurthy, Manoj Kumar, Hema A. Murthy, Xavier Serra
  ISMIR (2015)
  [Full Paper]

Teaching Assistant

• Speech Recognition and Processing for Multimedia at USC, Spring 2017 & Spring 2018


• Introduction to Electrical Engineering at IIT Madras, Fall 2014


• Advanced Communication Lab II at IIT Madras, Spring 2015

CV