Khaled Koutini

ML researcher, JKU, Linz, Austria

prof_pic.jpg

I am a postdoctoral researcher at Johannes Kepler University. I enjoy solving difficult problems and discovering new things. I am interested in machine learning, audio perception and understanding, and generative models. I worked on different machine learning topics, including architecture design, multi-modal learning and alignment, over-parameterization and scaling laws, model compression and quantization, domain adaptation, and adversarial robustness.

news

Jun 15, 2023 We achieved the first place in the DCASE 2023 challange in Language-Based Audio Retrieval and Low-Complexity Acoustic Scene Classification.
Jun 4, 2023 A paper is presented at ICASSP 2023: Efficient Large-scale Audio Tagging via Transformer-to-CNN Knowledge Distillation.
Nov 27, 2022 My PhD thesis is published: Inductive Bias in Learning General Audio Representations.
Sep 22, 2022 A paper is presented at INTERSPEECH 2022: Efficient Training of Audio Transformers with Patchout.

selected publications

2022

  1. Learning General Audio Representations With Large-Scale Training of Patchout Audio Transformers
    Khaled Koutini, Shahed Masoudian, Florian Schmid, Hamid Eghbal-zadeh, Jan Schlüter, and Gerhard Widmer
    NeurIPS challenge, Holistic Evaluation of Audio Representations (HEAR). Proceedings of Machine Learning Research, 2022
    PDF
  2. Efficient Training of Audio Transformers with Patchout
    Khaled Koutini, Jan Schlüter, Hamid Eghbal-zadeh, and Gerhard Widmer
    In Interspeech 2022, 23nd Annual Conference of the International Speech Communication Association, 2022
    HTML PDF

2021

  1. Receptive Field Regularization Techniques for Audio Classification and Tagging with Deep Convolutional Neural Networks
    Khaled Koutini, Hamid Eghbal-zadeh, and Gerhard Widmer
    IEEE/ACM Transactions on Audio, Speech, and Language Processing, 2021
    HTML PDF

2019

  1. The Receptive Field as a Regularizer in Deep Convolutional Neural Networks for Acoustic Scene Classification
    Khaled Koutini, Hamid Eghbal-zadeh, Matthias Dorfer, and Gerhard Widmer
    In 27th European Signal Processing Conference, EUSIPCO 2019, A Coruña, Spain, September 2-6, 2019, 2019
    HTML PDF