DeepFake Monitor project completed!

DeepFake Monitor is a project created by students of the Cybersecurity program at Wrocław Tech in collaboration with the company IDENTT. The coordinators of the students' work and collaborators with IDENTT were representatives of KN Solvro consisting of Wiktoria Mróz, Łukasz Lenkiewicz, Oleksii Furman, and Wojciech Cebula. 

Composition of the project teams from the Cybersecurity program: 
The two project teams included third-year undergraduate students from the Cybersecurity program:
Team I: Bartłomiej Baut, Konrad Bik, Konrad Kłoniecki, Krzysztof Mola, Joanna Turkiewicz, Jakub Ziontek
Team II: Marta Bołd, Bartosz Prysak, Marta Mikołajczyk, Aleksandra Tomaszek, Andżelika Korcz

Description

The goal of the "DeepFake Monitor" project was to create a system for monitoring social media to identify and collect data on DeepFake content. The project focused on actively searching for and documenting occurrences of DeepFakes. Participants in the project were tasked with collecting, cataloging, and analyzing data about DeepFakes, which could be used to improve existing AI tools designed to detect manipulated content. 

The students utilized a set of monitoring and advanced search tools to scan popular social media platforms (such as Facebook, Twitter, Instagram, TikTok) for potential DeepFakes. The teams conducted an analysis of various online sources where DeepFakes might appear. After selecting the sources, the teams created special web bots, or crawlers, that automatically searched these sites to find DeepFake materials. These crawlers collected data and forwarded it for further analysis. Each identified DeepFake was saved along with metadata such as the source, publication date, estimated reach, and a brief description of the content and context. This data was organized to facilitate further analysis. Based on the collected data, students prepared reports analyzing trends, techniques used in creating DeepFakes, and their impact on society.

Main action steps: 

• Social Media Monitoring 
• Data Collection and Cataloging
• Analysis and Reporting

Results:

• Building a valuable database of DeepFakes that will serve as an educational and research resource for scientists and developers working on AI technologies.
• Increasing public awareness about issues related to DeepFake content and promoting critical thinking in the context of social media. 
• Preparing students for active participation in the discourse on ethics and responsibility in the era of advanced image and video manipulation technologies.

Meeting at IDENTT's headquarters in Wrocław

To summarize the DeepFake Monitor project, Paweł Narolski, Head of Research and Development at IDENTT, invited the students and representatives of the Solvro Science Club to a meeting at the company's headquarters in Wrocław.  The meeting was also attended by the CEO - Jan Szajda. 

Students and representatives of KN Solvro shared their insights and reflections on the project, summarized their activities, and talked about the challenges they faced while working on the project. Both students and the project coordinators from KN Solvro received certificates and gadgets. At the end, the teams were given a tour of IDENTT's office.

Certificates for coordinators from KN Solvro:

Each project participant received a certificate confirming the skills acquired in the project.

Project Summary

Members of the Machine Learning section of KN Solvro managed the implementation of the DeepFake Monitor project, identifying threats related to generative artificial intelligence in social networks, striving to support and motivate the two student teams.

The students' reports were rated very highly! Congratulations to these two Cybersecurity teams from Wrocław Tech! 👏

Wiktoria, Łukasz, Oleksii, and Wojtek once again demonstrated their commitment, willingness, and showed how to lead effective collaboration on a technology project. Bravo! 💪

More about the project can be found on the portfolio page in the DeepFake Monitor project.