On Mitigating Acoustic Feedback in Hearing Aids with Frequency - - PowerPoint PPT Presentation

▶

Sep 08, 2022 437 likes •646 views

On Mitigating Acoustic Feedback in Hearing Aids with Frequency Warping by All-Pass Networks Presented at the 178 th Acoustical Society of America Meeting, December 2019 Harinath Garudadri hgarudadri@ucsd.edu +1 858 668 6128 Qualcomm Institute

SLIDE 1

On Mitigating Acoustic Feedback in Hearing Aids with Frequency Warping by All-Pass Networks Harinath Garudadri hgarudadri@ucsd.edu +1 858 668 6128 Qualcomm Institute of Calit2 University of California, San Diego

Presented at the 178th Acoustical Society of America Meeting, December 2019

SLIDE 2

2 / 16

To enable psychophysical investigations beyond what is possible today

– Miller and Donahue Open Speech Signal Processing Platform Workshop, NIH, Bethesda, MD, Oct. 2014

SLIDE 3

Pisha, et al., (2019), IEEE Access

SLIDE 4

4 / 16

BTE-RICs FPGA Multi-channel ePhys IMU FPGA Sensing & Actuation ALSA Custom FPGA IIO SPI Browser Enabled User Devices Command Line Interface libOSP Reference Designs libDSP NginX SQLite

Node.js /PHP

HTML Web Apps Laravel 802.15 PAN (BT) 802.11 WiFi xMAC

(experimental)

4G/5G

RT-MHA Embedded Web Server Communication

Linux Drivers

Wireless FPGA D C C B A E E E E F F

Processing and Communication Device (PCD) Transducer Devices (TDs)

OSP Development Model

ADB UART React H xPhy

(experimental)

GPS Daemon for syncing with AWS G A

Users and Clinical Researchers

Web Developers and DSP Engineers

C D

Signal Processing Researchers

Wireless & Comm Researchers Embedded Computing and Wearables

E G

Computer Scientists

SLIDE 5

5 / 16

BTE-RICs FPGA Multi-channel ePhys IMU FPGA Sensing & Actuation ALSA Custom FPGA IIO SPI Browser Enabled User Devices Command Line Interface libOSP Reference Designs libDSP NginX SQLite

Node.js /PHP

HTML Web Apps Laravel 802.15 PAN (BT) 802.11 WiFi xMAC

(experimental)

4G/5G

RT-MHA Embedded Web Server Communication

Linux Drivers

Wireless FPGA D C C B A E E E E F F

Processing and Communication Device (PCD) Transducer Devices (TDs)

OSP Development Model

ADB UART React H xPhy

(experimental)

GPS Daemon for syncing with AWS G A

Users and Clinical Researchers

Web Developers and DSP Engineers

C D

Signal Processing Researchers

Wireless & Comm Researchers Embedded Computing and Wearables

E G

Computer Scientists

SLIDE 6

Freping – A portmanteau for Frequency Warping

6/16

1 1−αz−1 (1−α2)z−1 1−αz−1 z−1−α 1−αz−1 z−1−α 1−αz−1

v(−n)

q0(n)

q1(n)

q2(n)

q3(n)

Allpass Network

Framing & windowing All-pass network (α) Overlap-add

Freping

Original signal Frequency- warped signal

Realtime frequency warping

Discrete Representation of Signals, Oppenheim and Johnson, IEEE Proceedings, 1972.

SLIDE 7

When do Hearing aids howl? Nyquist Stability Criteria (NSC) due to acoustic feedback

7/16

The class of LMS algorithms break the magnitude condition
Freping breaks both magnitude and phase conditions
G(ejω, n)

⇣ F(ejω, n) − ˆ F(ejω, n) ⌘

≥ 1,

(magnitude cond.) ∠G(ejω, n) ⇣ F(ejω, n) − ˆ F(ejω, n) ⌘ = m2π, (phase cond.) ˆ F(ejω, n) is the feedback path estimate.

SLIDE 8

Freping for AFC and Frequency Warping in RT-MHA

x(n)

+ +

d(n) Hearing Aid Processing G(z, n) e(n)

?

B(z) A(z, n) Copy of W(z, n) u(n)

−

ˆ y(n) AFC Filter W(z, n) uf(n) A(z, n)

+−

ˆ yf(n)

df(n) Coefficient Adaptation ef(n) Feedback Path F(z, n) y(n)

Freping

SLIDE 9

9 / 16

C-H Lee et al., Interspeech 2019

SLIDE 10

10 / 16

MPEG_es01_input.wav MPEG_es01_AFC.wav

SLIDE 11

11 / 16

MPEG_es01_FL.wav (frequency lowering in bands 4, 5) MPEG_es01_FU.wav (frequency increasing in bands 4, 5)

SLIDE 12

Machine Aided Self Fitting (Selfi) Research

Yeah, there’s an app for that!

12/16

{skin, skim}, {state, skate}, {peer, poor}, {lock, locks}, …

User preference (left) & Just Noticeable Differences

SLIDE 13

The Machine’s Role in Selfi Research

13/16

The structure of HA parameters is “known” – Closed form search techniques The structure of HA parameters is “unknown” – Stochastic search techniques

SLIDE 14

Big Data to Rescue

14/16

NHANES (~30,000 PTAs) → Clustering → NAL-NL2 prescriptions → Binary Search Tree Fitting (BSTFit) → Selfi Refinement

SLIDE 15

15 / 16

How to enable psychophysical investigations beyond what is possible today? – OSP

1. What discoveries can clinical researchers make with the platform? 2. What discoveries can we translate to clinical practice?

SLIDE 16

Takeaway Message

Researchers from multiple disciplines

– leverage contributions from others to advance their domain and

Participate in promoting hearing healthcare

16 / 16

Further details

Wednesday Morning (Crown)

– 3aPP3. Noise management features of the open speech platform – 3aPP4. Researcher and user interfaces for studies of hearing-aid self adjustment – 3aPP5. Open speech platform: Web-apps for hearing aids research

Wednesday Afternoon (Crown)

– 3pSP15. Self-fit generation of the wide range compression parameters in hearing aids

http://openspeechplatform.ucsd.edu/ and https://github.com/nihospr01/OpenSpeechPlatform-UCSD

SLIDE 17

Backup

SLIDE 18

Collaborators (2018)

SLIDE 19

19 / 34

SLIDE 20