Preloader

Call For Papers

Accepted Papers Proceedings on ACM Digital Lib
  • BuMA: Non-Intrusive Breathing Detection using Microphone Array
    Authors : Kaiyuan Hou, Stephen Xia, Xiaofan Jiang (Columbia University)
  • Beyond Microphone: mmWave-Based Interference-Resilient Voice Activity Detection
    Authors : Muhammed Zahid Ozturk (University of Maryland, College Park); Chenshu Wu (The University of Hong Kong; Beibei Wang (Origin Wireless Inc); Min Wu (University of Maryland, College Park); K.J.Ray Liu (Origin Wireless Inc.)
  • VoiceFind: Noise-Resilient Speech Recovery in Commodity Headphones
    Authors : Irtaza Shahid, Yang Bai, Nakul Garg, Nirupam Roy (University of Maryland, College Park)
  • Speech Disfluency Detection with Contextual Representation and Data Distillation
    Authors : Payal Mohapatra, Akash Pandey (Northwestern University); Bashima Islam (Worcester Polytechnic Institute); Qi Zhu (Northwestern University)
  • An Empirical Analysis of Perforated Audio Classification
    Authors : Mahathir Monjur, Shahriar Nirjon (UNC Chapel Hill)
  • Conversational AI Therapist for Daily Function Screening in Home Environments
    Authors : Jingping Nie (Columbia University); Hanya Shao (The Soho Center for Mental Health Counseling, Kensington Wellness); Minghui Zhao, Stephen Xia, Matthias Preindl, Xiaofan Jiang (Columbia University)
Scope and Topics

Among the five human senses, we most heavily rely on sound and vision in our everyday lives. Advances in computer vision have profoundly impacted our daily lives and greatly bridged the gap between human and machine intelligence. However, without exploiting the rich information about the activities, events, and environment around us embedded in audio, it is not possible to fully unleash the potential of machine intelligence. Though many physical characteristics of acoustics have been explored, it has been rarely explored with the lens of pervasive sensing modality in mobile and intelligent systems. Audio presents numerous opportunities, including – (1) reduced power consumption and computational requirements; (2) higher omnidirectional coverage; (3) more susceptibility to physical obstruction; (4) cheaper and smaller sensors; and (5) better privacy preservation. Acoustic intelligence will enable low-power, low-cost, real-time, hands-free, and more interactive communication between machines, humans, and the environment. This workshop aims to push the boundaries of machine intelligence by exploring the sensing, communication, system implementation, algorithm design, and applications of intelligent acoustic systems.


IASA welcomes contributions in all aspects of acoustic sensing, systems, and applications, including – mobile and wearable sensing or communication utilizing sound, vibration, ultrasound, and infrasound, algorithms in acoustic intelligence, security, and privacy utilizing acoustics, and data or deployment experiences.


Topics of interest include, but not limited to:
  • Acoustic mobile applications and embedded systems
  • Novel acoustic sensors
  • Innovative acoustic sensing wearables, devices, and platforms
  • Novel software architectures for audio processing
  • Authentication and verification with audio
  • Privacy preservation of continuous listening devices
  • New acoustic features for information extraction
  • Audio augmented reality, AR/VR/Immersive reality applications of acoustic
  • Audio synthesis
  • Sound localization, denoising, and separation
  • Resource-efficient machine learning and artificial intelligence using acoustic signals
  • Multi-modal sensing and learning where acoustic modality enhances or gets enhanced by other sensing modalities
  • Audio + X; where X = edge computing, localization, battery-less, deep learning, robotics, etc
  • Multi-task learning leveraging wearables with microphones
  • Acoustic passive sensing support for vehicular and mobile robotic systems
  • Acoustic sensing in miniaturized aerial devices
  • Communication with audible sounds, ultra-sounds, infra-sound, and vibration
  • Quality-aware audio data collection with earables and wearables
  • Health and wellbeing applications utilizing acoustic signals
  • Emerging applications in earables based on audio signals

In addition to the research papers, we also invite “Challenge Papers” that present revolutionary new ideas that challenge existing assumptions prevalent among acoustic signal processing, machine learning, and the mobile systems community in general. These “challenge papers” should provide stimulating ideas that may open up exciting avenues and/or influence the direction of future intelligent acoustic research. An exhaustive evaluation of the proposed ideas is not necessary for this category; instead, insight and in-depth understanding of the issues are expected.


Submission Guideline

We invite original research papers that have not been previously published and are not currently under review for publication elsewhere. Submitted papers should be no longer than six pages for research papers and four pages for challenge papers (including references and appendices). Your submission must use a 10pt font (or larger) and be correctly formatted for printing on Letter-sized (8.5" by 11") paper. Paper text blocks must follow ACM guidelines: double-column, with each column 9.25" by 3.33", 0.33" space between columns and single-spaced. The title of challenge papers must bear a "Challenge:" prefix.


Papers are to be submitted at https://iasa2022.hotcrp.com


All accepted papers will be published as part of the ACM proceedings.


Important Dates
  • Paper submission: May 6, 2022, AoE
  • Notification of Acceptance: May 16, 2022, AoE
  • Camera Ready Deadline: May 28, 2022, AoE
  • Workshop Date: July 1, 2022
  • Location: Portland, Oregon, USA

Workshop committees

General co-Chairs

Xiaofan (Fred) Jiang (Columbia University)
Nirupam Roy (University of Maryland, College Park)
Shahriar Nirjon (University of North Carolina, Chapel Hill)


TPC co-Chairs

Chulhong Min (Nokia Bell Labs)
Stephen Xia (Columbia University)
Bashima Islam (Worcester Polytechnic Institute)


Steering Committee


Technical Program Committee


Program

Local Time: Pacific Standard Time(GMT-8)

8:15 am - 8:30 am Welcome note from the organizers
8:30 am - 9:30 am Keynote 1: Wontak Kim ( Amazon 126 )
Title: From Smart to Smarter - Evolution of Smart Speakers in the Age of Ambient Sensing and Computing
9:30 am - 10:30 am Session 1: Audio in Health
1. BuMA: Non-Intrusive Breathing Detection using Microphone Array
2. Speech Disfluency Detection with Contextual Representation and Data Distillation
3. Conversational AI Therapist for Daily Function Screening in Home Environments
10:30 am - 11:00 am Coffee Break
11:00 am - 12:00 pm Session 2: Hear Better
1. Beyond Microphone: mmWave-Based Interference-Resilient Voice Activity Detection
2. VoiceFind: Noise-Resilient Speech Recovery in Commodity Headphones
3. An Empirical Analysis of Perforated Audio Classification
12:00 pm - 12:30 pm Invited Talks: Nivedita Arora
Title: Self-powered Acoustic Vibration Sensing Material
12:30 pm - 1:30 pm Lunch
1:30 pm - 2:30 pm Keynote 2: Romit Roy Choudhury ( University of Illinois at Urbana-Champaign )
Title: Some Challenging Problems in Earable Computing
2:30 pm - 3:30 pm Panel Discussion:
  • - Shyam Gollakota ( University of Washington )
  • - VP Nguyen ( The University of Texas at Arlington )
  • - Shijia Pan ( University of California, Merced )
Topic: Intelligent acoustics: Why Now? What's Next?
3:30 pm - 4:00 pm Coffee Break
4:00 pm - 4:30 pm Business Meeting
4:30 pm - 4:45 pm Closing Remarks

Talks

Speakers

Speaker

Wontak Kim

Manager, Audio Research Science. Amazon Lab126

Title: From Smart to Smarter - Evolution of Smart Speakers in the Age of Ambient Sensing and Computing


Echo and Alexa started the era of smart speakers paving a way to voice centric interaction with machine intelligence. In this talk, we will present the technical evolution in the creation of smart speakers and also touch upon remaining challenges and opportunities for the future. In the beginning of smart speaker experience, the typical use cases were on playing music, checking weather, fetching news, setting alarms etc. The technologies that enabled those were: farfield voice pickup, speech recognition, key word spotting, natural language processing, text to speech, local/cloud hybrid compute engine to name a few. Lately they evolved to become more than an audio device - equipped with screens and cameras, it features video calls and other screen based functionalities. In addition they are becoming the key building blocks in home automation. It does not only function as the main user interface but also equipped with different sensing mechanisms - acoustic (audible and ultrasonic), CV, wifi, temperature and light. In the mean time Echo devices as a group can now be used as smart home security system detecting fire alarm, house intrusion, and human presence. With the addition of Astro and Ring drone camera, home surveillance capability is even more sophisticated. Looking ahead, we see a tremendous possibility in smart speakers as ambient sensing and compute system enabling the machine intelligence to be always on and available yet not intrusive to our daily lives. It will ideally anticipate our needs and serve the right information or actions the right time.

Speaker

Romit Roy Choudhury

University of Illinois at Urbana-Champaign

Title: Some Challenging Problems in Earable Computing


Earable computing refers to the possibility that tomorrow’s earphones and glasses will become more that just accessories of the smartphone, offering a host of new capabilities and applications. The wheels are already in motion and solutions to some (difficult) research problems could even unlock technological disruptions. In this talk, I plan to discuss a few research problems in this space, including source separation, speech enhancement, spatial audio, motion tracking, and human-feedback based non-convex optimization.

Organizers



Organizers
Shahriar Nirjon

University of North Carolina

Shahriar Nirjon

University of North Carolina
Organizers
Xiaofan (Fred) Jiang

Columbia University

Xiaofan (Fred) Jiang

Columbia University
Organizers
Nirupam Roy

UMD College Park

Nirupam Roy

UMD College Park