AIDA Course: “An Introduction to Trustworthy Machine Learning”, 23-24/11/2022

November 16th, 2022 Daniela Lopez de Luise

Idiap Research Institute organizes an online AIDA short course on Trustworthy ML offered through the International Artificial Intelligence Doctoral Academy (AIDA). 

 

The purpose of this course is to overview the foundations and the current state-of-the-art in differentially private ML and adversarial examples for privacy protection.

 

This short course will cover the following topics:

  • Introduction and motivation (privacy and personal data)
  • Differential privacy

1.       Definitions

1.       Differentially private machine learning

  • Adversarial examples

1.       Adversarial goals, knowledge, and properties

1.       Defenses against adversarial examples

1.       Norm-bounded and content-based adversarial examples

  • Adversarial examples for privacy protection

1.       Privacy and utility for images/audio in social multimedia

  • Hands-on examples (with software modules distributed to the participants)

 

 

LECTURER: 

– Sina Sajadmanesh, email: sajadmanesh@idiap.ch

– Ali Shahin Shamsabadi, email: a.shahinshamsabadi@turing.ac.uk

– Daniel Gatica Perez, email: gatica@idiap.ch

 

HOST INSTITUTION/ORGANIZER: IDIAP

 

REGISTRATION: Free of charge.

 

WHEN: 23-24 November 2022 from 10:00 to 13:00 CET

WHERE: Online via Zoom

 

HOW TO REGISTER and ENROLL: 

Both AIDA and non-AIDA students are encouraged to participate in this short course.

 

If you are an AIDA Student* already, please:

Step (a) register for the course by sending an email to sajadmanesh@idiap.ch, AND

Step (b) enroll in the same course in the AIDA system using the enrollment button on the AIDA course page, so that this course enters your AIDA Course Attendance Certificate.

 

If you are not an AIDA Student do only step (a).

 

*AIDA Students should have been registered in the AIDA system already (they are PhD students or PostDocs that belong only to the AIDA Members listed on this page:

https://www.i-aida.org/about/members/)

 

Sina Sajadmanesh

Email sajadmanesh@idiap.ch

 

 

More details: https://www.i-aida.org/course/an-introduction-to-trustworthy-machine-learning/

 

Read More »

CfP JIVP SI on “Manipulation Detection in Digital Images and Videos”

November 16th, 2022 Daniela Lopez de Luise
Call for papers
Final Extension!

Working on Manipulation Detection in Digital Images and Videos?

-> We are organizing a Special Issue at EURASIP Journal on Image and Video Processing!

Link: https://jivp-eurasipjournals.springeropen.com/manipulation-detection-in-digital-images-and-videos

Submission deadline: 15 December 2022

Guest Editors: Antitza Dantcheva, Abhijit Das, Hu Han, Christian Rathgeb, Naser Damer, Luisa Verdoliva, Ruben Tolosana

Read More »

Call for Papers – ICMR 2023 – International Conference on Multimedia Retrieval

November 16th, 2022 Daniela Lopez de Luise
ACM International Conference on Multimedia Retrieval 2023

Thessaloniki, Greece, 12 – 15 June 2023

Web: https://icmr2023.org/

**********************************************************
_______________

CALL FOR PAPERS
_______________

ACM ICMR 2023 is calling for high quality original papers addressing
innovative research in multimedia retrieval and its related broad
fields. The main scope of the conference is not only search and
retrieval of multimedia data but also analysis and understanding of
multimedia contents including community-contributed social data,
lifelogging data and automatically generated sensor data, integration of
diverse multimodal data, deep learning-based methodology and practical
multimedia applications.

Long research papers should present complete work with evaluations on
topics related to the Conference. They will have both oral and poster
presentations at the conference. Authors of the best papers will be
offered an opportunity to extend their work for a Special Issue in a
peer-reviewed multimedia journal (to be defined). Short research papers
should present preliminary results or more focused contributions. They
will be presented as posters at the conference.

Topics of Interest

     -Multimedia content-based search and retrieval,
     -Multimedia-content-based (or hybrid) recommender systems,
     -Large-scale and Web-scale multimedia retrieval,
     -Multimedia content extraction, analysis, and indexing,
     -Multimedia analytics and knowledge discovery,
     -Multimedia machine learning, deep learning, and neural networks,
     -Relevance feedback, active learning, and transfer learning,
     -Fine-grained retrieval for multimedia,
     -Event-based indexing and multimedia understanding,
     -Semantic descriptors and novel high- or mid-level features,
     -Crowdsourcing, community contributions, and social multimedia,
     -Multimedia retrieval leveraging quality, production cues, style,
framing, and affect,
     -Synthetic media generation and detection,
     -Narrative generation and narrative analysis,
     -User intent and human perception in multimedia retrieval,
     -Query processing and relevance feedback,
     -Multimedia browsing, summarization, and visualization,
     -Multimedia beyond video, including 3D data and sensor data,
     -Mobile multimedia browsing and search,
     -Multimedia analysis/search acceleration, e.g., GPU, FPGA,
     -Benchmarks and evaluation methodologies for multimedia
analysis/search,
     -Privacy-aware multimedia retrieval methods and systems,
     -Fairness and explainability in multimedia analysis/search,
     -Legal, ethical and societal impact of multimedia retrieval research,
     -Applications of multimedia retrieval, e.g., news/journalism,
media, medicine, sports, commerce, lifelogs, travel, security, and
environment.

_____________________

SUBMISSION GUIDELINES
_____________________

Maximum Length of a Paper

Long research paper: Each long research paper should not be longer than
8 pages, plus additional pages for the list of references.

Short research paper: Each short research paper should not be longer
than 4 pages, plus additional pages for the list of references.

_______________

IMPORTANT DATES
_______________

Paper Submission Due: January 31, 2023
Notification of Acceptance: March 31, 2023
Camera-Ready Papers Due: April 17, 2023

______________

REVIEW PROCESS
______________

ACM ICMR follows a double-blind review process for full paper selection.
Authors should not know the names of the reviewers of their papers, and
reviewers should not know the name(s) of the author(s). Please prepare
your paper in a way that preserves anonymity of the authors:

     Do not put your names under the title,
     Avoid using phrases such as “our previous work” when referring to
earlier publications by the authors,
     Remove information that may identify the authors in the
acknowledgments (e.g., co-workers and grant IDs),
     Check supplemental material for information that may identify the
authors’ identity,
     Avoid providing links to Websites that identify the authors.

Abstract and Keywords

The abstract and the keywords form the primary source for assigning
papers to reviewers. So make sure that they form a concise and complete
summary of your paper with sufficient information to let someone who has
not read the full paper know what it is about.

Submission Instructions: https://icmr2023.org/paper-submissions/

_______

CONTACT
_______

For any question regarding full and short paper submissions, please
visit the conference website (icmr2023.org) or email the Program Chairs:

     Vasileios Mezaris, Centre for Research and Technology Hellas,
Greece (bmezaris@iti.gr)
     Symeon Papadopoulos, Centre for Research and Technology Hellas,
Greece (papadop@iti.gr)
     Adrian Popescu, CEA LIST, France (adrian.popescu@cea.fr)
     Zi (Helen) Huang, University of Queensland, Australia
(huang@itee.uq.edu.au)

Read More »

CALL FOR BOOK CHAPTER (Adversarial Multimedia Forensics)

November 16th, 2022 Daniela Lopez de Luise

We are pleased to invite you to submit a chapter for inclusion in the “Adversarial Multimedia Forensics” book that will be published by Springer – Advances in

Information Security. Chapter submissions should be 15 to 20 pages long, single-spaced, and single-column in latex, and should provide enough information

for readers and professionals in Cybersecurity Applications, particularly with regard to multimedia forensics and security. Multimedia forensics, counter-forensics,

and anti-counter-forensics constitute the three primary sections of the book.

Looking For: We are seeking for chapters that apply security and attack principles, methodologies, and strategies to the multimedia domain i.e., images and

videos as given in the call. As an example, from multimedia forensics, consider source camera identification or fine-grained CFA artifacts assessment for forgery

detection. Using adversarial attacks on medical images or considering the transferability of adversarial attacks on vision transformers are other examples of counterforensics.

The adversary-aware double JPEG-detector via selective training on attacked samples and JPEG compression image contrast manipulation identification

are indeed a few examples of Anti-Counter Forensics against exploratory attacks. In contrast, examples employing Anti-Counter Forensics against causative attacks

would include proposing defenses against poisoning attacks to satellite imagery models. Machine Learning (ML) is becoming the de-facto standard for Multimedia

Forensics (MF) due to its exceptional capabilities. However, the peculiarity of ML architectures leads to new, significant security vulnerabilities that prohibit their

usage in security-critical applications like MF, where it is inconceivable to ignore the potential existence of an adversary. However, given the weakness of the traces

that forensic techniques rely on, disabling forensic analysis turns out to be a simple task. Determining the security of ML-based systems in the presence of an

adversary and developing innovative strategies capable of improving their protection are therefore of utmost relevance. In order to overcome the security constraints

of ML models used as counter-forensics techniques, it has become crucial for MF to develop adversary-aware solutions. This book contributes to the aforementioned

goal by emphasizing on image manipulation detection using ML/DL algorithms for MF in adversarial environments. The main structure of the book is divided into the

following three sections: (I) presents different methodologies in multimedia forensics; (II) and discusses general concepts and terminology in the field of

adversarial machine learning (Adv-ML), with a focus on the concern of counter-forensics (CF), and anti-counter forensics.

Originality: Chapter contributions should contain 25-30% novel content compared to earlier published work by the authors.

Submission: There are no submission or acceptance fees for manuscripts submitted to this book for publication. All manuscripts are accepted based on a

double-blind peer review editorial process. Please send your manuscript *.pdf, *.tex to the e-mail address of one of the editors (ehsan.nowroozi@eng.bau.edu.tr,

Alireza.jolfaei@flinders.edu.au, Kassem.kallas@inria.fr)



Timeline: Expression of interest: 15-Jan-2023 (tentative: chapter title, and abstract): Send by email to editors, Selection of chapters: 30-Jan-2023

(Inform to Authors by email and share Easy Chair link with authors for submission), Deadline for full chapter submission: 30-Feb-2023 (Submit via

Easy Chair), Review of chapters: 30-Mar-2023, Camera-ready version: 20-April-2023 (Submit in Easy chair)

Book Areas

As we mentioned, the core of the book consists of: (I) Multimedia Forensics, (II) Counter-Forensics, and (III) Anti-Counter-Forensics. The tentative table of

contents will be:

(Part-I) Multimedia Forensics: This section discusses machine learning and deep learning techniques for digital image forensics and image tampering detection.

Recent forensic analysis techniques will be covered in this part, including (I) acquisition-based footprints, (II) coding- based footprints, and (III) editing-based

footprints.

(Part-II) Counter-Forensics: This section will explain counter-forensics (CF), which is the counterpart of the detector and refers to any techniques intended to thwart

a forensic investigation. This is also referred to as anti-forensics in the literature. Deep learning and adversarial attacks on a machine learning model in this case can

be divided into exploratory and causative. This part discusses the various methods that have been proposed so far to mitigate a forensic analysis.

Exploratory Attacks: The exploratory attack scenario restricts the adversary's ability to changes to test data and forbids changes to training examples.

Example 1: Adversarial Cross-Modal Attacks from Images to Videos, Example 2: Adversarial attacks on a medical images.

Causative attacks: In causative attacks, the offensive can disrupt the training process to inject a backdoor into the model to be exploited later at inference

time; these attacks are commonly referred to as poisoning, backdoor or Trojan attacks, Example 1: Attacks using backdoors against Vision Transformers,

Example 2: Performing Backdoor Attacks Using Rotation Transformation

(Part-III) Anti Counter-Forensics: To protect the reliability of the forensic analysis, numerous anti- CF techniques have been developed in response to CF. The

majority of these methods are appropriate for particular CF methods. This section will explain recent advances in anti counter-forensics methods.

Defense against Exploratory Attacks: This part will discuss recent methods that have been proposed so far for improving the security of detectors against

exploratory attacks, such as adversary- aware detectors and developing a secure architectures, Example 1: Adversary-Aware Double JPEG-Detector via

Selected Training on Attacked Samples, Example 2: Resistant to JPEG Compression Image Contrast Manipulation Identification.

Defense against Causative Attacks: In this section, we will survey the recent techniques that have been proposed so far for enhancing the security of model

against poisoning attacks, Example 1: Defense against poisoning attacks on satellite imagery models, Example 2: Using Heatmap Clustering to find Deep

Neural Network Backdoor Poisoning Attacks



Book Editors

Dr. Ehsan Nowroozi, Assistant Professor, Bahcesehir University, Istanbul, Turkey (ehsan.nowroozi@eng.bau.edu.tr)

Dr. Alireza Jolfaei, Associate Professor, Flinders University, Adelaide, Australia (Alireza.jolfaei@flinders.edu.au)

Dr. Kassem Kallas, Research Scientist at INRIA, Rennes, France (Kassem.kallas@inria.fr)

Advertisement: https://enowroozi.com/call-for-book-chapter-adversarial-multimedia-forensics/

Read More »

DeepLearn 2023 Spring: early registration November 13

November 15th, 2022 Daniela Lopez de Luise
9th INTERNATIONAL SCHOOL ON DEEP LEARNING

DeepLearn 2023 Spring

Bari, Italy

April 3-7, 2023

https://irdta.eu/deeplearn/2023sp/

***********

Co-organized by:

Department of Computer Science
University of Bari “Aldo Moro”

Institute for Research Development, Training and Advice – IRDTA
Brussels/London

******************************************************************

Early registration: November 13, 2022

******************************************************************

SCOPE:

DeepLearn 2023 Spring will be a research training event with a global scope aiming at updating participants on the most recent advances in the critical and fast developing area of deep learning. Previous events were held in Bilbao, Genova, Warsaw, Las Palmas de Gran Canaria, Guimarães, Las Palmas de Gran Canaria, Luleå and Bournemouth.

Deep learning is a branch of artificial intelligence covering a spectrum of current exciting research and industrial innovation that provides more efficient algorithms to deal with large-scale data in a huge variety of environments: computer vision, neurosciences, speech recognition, language processing, human-computer interaction, drug discovery, health informatics, medical image analysis, recommender systems, advertising, fraud detection, robotics, games, finance, biotechnology, physics experiments, biometrics, communications, climate sciences, bioinformatics, geographic information systems, etc. etc. Renowned academics and industry pioneers will lecture and share their views with the audience.

Most deep learning subareas will be displayed, and main challenges identified through 24 four-hour and a half courses and 3 keynote lectures, which will tackle the most active and promising topics. The organizers are convinced that outstanding speakers will attract the brightest and most motivated students. Face to face interaction and networking will be main ingredients of the event. It will be also possible to fully participate in vivo remotely.

An open session will give participants the opportunity to present their own work in progress in 5 minutes. Moreover, there will be two special sessions with industrial and recruitment profiles.

ADDRESSED TO:

Graduate students, postgraduate students and industry practitioners will be typical profiles of participants. However, there are no formal pre-requisites for attendance in terms of academic degrees, so people less or more advanced in their career will be welcome as well. Since there will be a variety of levels, specific knowledge background may be assumed for some of the courses. Overall, DeepLearn 2023 Spring is addressed to students, researchers and practitioners who want to keep themselves updated about recent developments and future trends. All will surely find it fruitful to listen to and discuss with major researchers, industry leaders and innovators.

VENUE:

DeepLearn 2023 Spring will take place in Bari, an important economic centre on the Adriatic Sea. The venue will be:

University of Bari “Aldo Moro”

STRUCTURE:

3 courses will run in parallel during the whole event. Participants will be able to freely choose the courses they wish to attend as well as to move from one to another.

Full live online participation will be possible. However, the organizers highlight the importance of face to face interaction and networking in this kind of research training event.

KEYNOTE SPEAKERS: (to be completed)

Vipin Kumar (University of Minnesota), Knowledge Guided Deep Learning: A Framework for Accelerating Scientific Discovery

William S. Noble (University of Washington), Deep Learning Applications in Mass Spectrometry Proteomics and Single-Cell Genomics

PROFESSORS AND COURSES: (to be completed)

Patrick Gallinari (Sorbonne University), [intermediate] Physics Aware Deep Learning for Modeling Dynamical Systems

Sergei V. Gleyzer (University of Alabama), tba

Jacob Goldberger (Bar-Ilan University), [introductory/intermediate] Latent Random Variables, Generative Models and Variational Autoencoders

Yingbin Liang (Ohio State University), [intermediate/advanced] Bilevel Optimization and Applications in Deep Learning

Xiaoming Liu (Michigan State University), [intermediate] Deep Learning for Trustworthy Biometrics

Michael Mahoney (University of California Berkeley), tba

Razvan Pascanu (DeepMind), [intermediate] Understanding Learning Dynamics in Deep Learning and Deep Reinforcement Learning

Bhiksha Raj (Carnegie Mellon University), [introductory] An Introduction to Quantum Neural Networks [with Rita Singh and Daniel Justice]

Bart ter Haar Romeny (Eindhoven University of Technology), [intermediate/advanced] Explainable AI from First Principles

Martin Schultz (Research Centre Jülich), [introductory/intermediate] Deep Learning for Air Quality, Weather and Climate

Adi Laurentiu Tarca (Wayne State University), [intermediate] Machine Learning for Cross-Sectional and Longitudinal Omics Studies

Zhi Tian (George Mason University), [intermediate] Communication-Efficient and Robust Distributed Learning

Michalis Vazirgiannis (Polytechnic Institute of Paris), [intermediate/advanced] Graph Machine Learning with GNNs and Applications

Atlas Wang (University of Texas Austin), [intermediate] Sparse Neural Networks: From Practice to Theory

Guo-Wei Wei (Michigan State University), [introductory/advanced] Discovering the Mechanisms of SARS-CoV-2 Evolution and Transmission

Xiaowei Xu (University of Arkansas Little Rock), [intermediate/advanced] Deep Learning Language Models and Causal Inference

OPEN SESSION:

An open session will collect 5-minute voluntary presentations of work in progress by participants. They should submit a half-page abstract containing the title, authors, and summary of the research to david@irdta.eu by March 26, 2023.

INDUSTRIAL SESSION:

A session will be devoted to 10-minute demonstrations of practical applications of deep learning in industry. Companies interested in contributing are welcome to submit a 1-page abstract containing the program of the demonstration and the logistics needed. People in charge of the demonstration must register for the event. Expressions of interest have to be submitted to david@irdta.eu by March 26, 2023.

EMPLOYER SESSION:

Organizations searching for personnel well skilled in deep learning will have a space reserved for one-to-one contacts. It is recommended to produce a 1-page .pdf leaflet with a brief description of the company and the profiles looked for to be circulated among the participants prior to the event. People in charge of the search must register for the event. Expressions of interest have to be submitted to david@irdta.eu by March 26, 2023.

ORGANIZING COMMITTEE:

Donato Malerba (Bari, local chair)
Carlos Martín-Vide (Tarragona, program chair)
Sara Morales (Brussels)
David Silva (London, organization chair)

REGISTRATION:

It has to be done at

https://irdta.eu/deeplearn/2023sp/registration/

The selection of 8 courses requested in the registration template is only tentative and non-binding. For the sake of organization, it will be helpful to have an estimation of the respective demand for each course. During the event, participants will be free to attend the courses they wish.

Since the capacity of the venue is limited, registration requests will be processed on a first come first served basis. The registration period will be closed and the on-line registration tool disabled when the capacity of the venue will have got exhausted. It is highly recommended to register prior to the event.

FEES:

Fees comprise access to all courses and lunches. There are several early registration deadlines. Fees depend on the registration deadline. The fees for on site and for online participation are the same.

ACCOMMODATION:

Accommodation suggestions will be available in due time at

https://irdta.eu/deeplearn/2023sp/accommodation/

CERTIFICATE:

A certificate of successful participation in the event will be delivered indicating the number of hours of lectures.

QUESTIONS AND FURTHER INFORMATION:

david@irdta.eu

ACKNOWLEDGMENTS:

University of Bari “Aldo Moro”

Rovira i Virgili University

Institute for Research Development, Training and Advice – IRDTA, Brussels/London

Read More »
« Older Entries
Newer Entries »
Design by 2b Consult