Connectionists: Call for papers: Submission to special session proposal in ICASSP 2024

[Zhao Ren]

Dear communities,

we are recently planning to organise a special session in ICASSP 2024  
which will be in Seoul, Korea (https://2024.ieeeicassp.org).
  If the you are interested to submit a paper, we will include the  
information from the potential authors in our proposal,  
including TITLE, AUTHORS, AFFILIATIONS, AND SHORT ABSTRACTS.

As the submission deadline of the proposal is 28 June 2023 and only a  
few abstracts can be further included, it would be great to get  
contact with me about your interest by 23 JUNE 2023. The abstract will  
be expected by 28 JUNE 2023.

You can find more information about the special session:
**********************
SPECIAL SESSION: AI SECURITY AND PRIVACY IN BIOSIGNAL PROCESSING FOR  
HEALTHCARe
Security and privacy have become increasingly important, as artificial  
intelligence (AI) begins to influence our daily lives. To train robust  
models and make accurate predictions, machine learning models'  
security levels are essential to be enhanced. Recent research has  
shown that deep learning models are susceptible to external attacks  
including adversarial attacks and spoofing attacks. Spoofing attacks  
deceive AI systems by pretending to be someone else, whereas  
adversarial attacks produce hostile samples that humans cannot tell  
apart from actual data. The issue is how to increase the model's  
defences against attacks. Additionally, it is crucial to uphold data  
protection laws like the General Data Protection Regulation (GDPR), in  
order to preserve users’ privacy and personal information. For  
instance, it is challenging to protect users’ privacy when uploading  
personal data to cloud servers. 


The processing of biosignals, such as acoustic signals, heart rates,  
and brain and muscle activities, has been investigated and  
successfully used in the healthcare domain for several decades.  
Numerous research studies on biosignal processing have incorporated  
AI, particularly deep learning, with great success. Security and  
privacy of AI systems are two crucial challenges in the healthcare  
area due to the highly sensitive nature of health-related data and the  
increasing demand for robust, trustworthy, and reliable AI systems for  
healthcare applications. 


    


   With this special session, we expect to attract high-quality  
research works in AI security and privacy in biosignal processing for  
healthcare. The topics include, but are not limited to: 


   - Federated learning for processing multi-client biosignal data  

   - Generating adversarial biosignal data for deceiving AI models 

   - Algorithms for protecting AI models against attacks  

   - Machine unlearning for forgetting users’ data 

   - Exploration of explainable AI in improving interpretability 

Session organisers:
Zhao Ren, University of Bremen, Germany
Björn W. Schuller, Imperial College Lodon, UK
Tanja Schultz, University of Bremen, Germany
************************


Thank you very much and best regards,

Zhao
 
  Dr.-Ing. Zhao Ren
Cognitive Systems Lab (CSL)
University of Bremen
Germany
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20230622/9baf6f42/attachment.html>