8:30 – 9:00

Introductions and Welcome

Steve Willenborg (VP Sales, Linkplay – APDA Chair)

The continued development of improved audio products relies on new techniques and development tools. At ADPA we strive to showcase the technologies and methods that are essential to product development engineering using the most advanced tools available including many that will be presented today. The use of DSP and AI/ML will continue to influence and enable new and exciting audio designs. These technologies are rapidly evolving and with the combination of deep learning and electro-acoustic engineering it is a very exciting time in the audio/acoustics products field.

9:00 – 9:50

Overcoming the Challenges of High-Quality Conferencing

Paul Beckmann (Chief Technical Officer, DSP Concepts)

Voice conferencing products are notoriously difficult to build.   Even relatively basic designs such as a single mic “voice puck” requires hours of testing and tuning to achieve a satisfying user experience.  More advanced designs with microphone arrays and satellite microphones introduce finicky, potentially time consuming problem-solving at both the hardware and software levels.  Meeting the rigorous standards for Microsoft Teams and Zoom certification requires significant expertise and specialized facilities.  In this talk, DSP Concepts CTO Paul Beckmann discusses the challenges in making high-quality conferencing and the key steps to achieving a differentiated product.

09:50-10:00

Break

10:00 – 11:30

Neural Network Design for Voice on the Edge

Josh Morris (Senior Manager, ML Development, DSP Concepts)

Learn about neural network design for always-on streaming audio classification for the edge. We’ll cover important topics including feature selection, layer selection, and tips and tricks for shrinking a model size. We’ll show real world use cases including keyword spotting and profiling using Audio Weaver and Tensorflow.

11:30 – 13:00

Lunch – Buffet

Demos and networking

Table demonstrations

Customer Products

Acer SK130 conference speaker

LG GP9 gaming speaker

               TalkTogether Full-Duplex Comparison Demo

               R-TASC Common Issues Presentation

               RAPID Kit with IntelliGo MLNR

13:00 – 14:00

The Evolving Use Cases for Video Conferencing

Eric Skup (VP Customer Acquisition and Success, DSP Concepts)

Michael Halloran (Sales Engineer, DSP Concepts)

Video conferencing has become an essential communication tool for businesses, educational institutions, and individuals worldwide. The rise of remote work and online learning has made video conferencing essential to conduct meetings, interviews, webinars, and classes. As the technology advances, new use cases are emerging such as telemedicine, virtual events, remote collaboration, augmented reality meetings, and watch parties.

Built with Audio Weaver, the TalkTogether: Virtual Co-Viewing solution elevates watch party experiences for streaming video with remarkably robust ML noise reduction. Even against loud action movie soundtracks, VCV delivers clear, full duplex conversation. Attendees will learn about the key elements in the processing chain followed by a live demonstration that steps through the signal flow and shows performance for distant far-field use cases.

14:00 – 15:30

Audio Weaver Training Part 1

Michael Halloran (Sales Engineer, DSP Concepts)

Audio Weaver Training – 1

Discover how Audio Weaver makes audio innovation easy in this introductory training session.  This short session is a preview of the full online training offered once per month by DSP Concepts.

Audio Weaver Overview

Audio Weaver Designer GUI

Standard Designer Processing Modules

Audio Weaver Core Embedded Libraries

Audio Weaver Use Cases

15:30 – 16:00

Break

16:00 – 17:30

Audio Weaver Training Part 2

Michael Halloran (Sales Engineer, DSP Concepts)

Audio Weaver Platform Training – 2

Part 2 of the Audio Weaver training session provides hands-on experience with Audio Weaver centered on a conferencing use case. Participants will learn how to navigate the TalkTogether voice comms layout and tune playback running on the STMicro STM32H747 MCU.