How To Apply Deep Learning for Signals and Sound

Deep Learning for Signals and Sound

Presented By Emelie Andersson and Johanna Pingel
Emelie Andersson and Johanna Pingel
Emelie Andersson and Johanna Pingel

We will demonstrate deep learning to denoise speech signals and generate musical tunes. You will see how you can use MATLAB to: 

  • Train neural networks from scratch using LSTM and CNN network architectures

  • Use spectrograms and wavelets to create 3d representations of signals

  • Access, explore, and manipulate large amounts of data

  • Use GPUs to train neural networks faster

Bios:

Emelie Andersson is an application engineer at MathWorks focusing on MATLAB applications such as data analytics, machine learning and deep learning. In her role she supports customers to adapt MATLAB products in the entire data analytics workflow. She has been with MathWorks for 2 years and holds an M.Sc. degree from Lund University in image analysis and signal processing.

Johanna Pingel joined the MathWorks team in 2013, specializing in Image Processing and Computer Vision applications with MATLAB. She has an M.S. degree from Rensselaer Polytechnic Institute and a B.A. degree from Carnegie Mellon University. She has been working in the Computer Vision application space for over 5 years, with a focus on object detection and tracking.

Presentation Description

Deep learning networks are proving to be versatile tools. Originally intended for image classification, they are increasingly being applied to a wide variety of other data types. In this webinar, we will explore deep learning fundamentals which provide the basis to understand and use deep neural networks for signal data. Through two examples, you will see deep learning in action, providing the ability to perform complex analyses of large data sets without being a domain expert. 

Explore how MATLAB addresses the common challenges encountered using CNNs and LSTMs to create systems for signals and sound and see new capabilities for deep learning for signal data.

Highlights:

We will demonstrate deep learning to denoise speech signals and generate musical tunes. You will see how you can use MATLAB to: 

  • Train neural networks from scratch using LSTM and CNN network architectures

  • Use spectrograms and wavelets to create 3d representations of signals

  • Access, explore, and manipulate large amounts of data

  • Use GPUs to train neural networks faster

Presentation Curriculum

Deep Learning for Signals and Sound
29:45
Hide Content