On January 30, 2021, the 21st-1st online workshop on CSIG multimedia forensics and security was held, hosted by China Society of Image and Graphics(CSIG), IEEE Beijing Section Biometrics Council Chapter and organized by the CSIG Technical Committee on Digital Media Forensics and Security and the Center for Research on Intelligent Perception and Computing of the Institute of Automation, Chinese Academy of Sciences. Professor Siwei Lyu from University at Buffalo, State University of New York and Dr.Ping Li from the Hong Kong Polytechnic University were invited to conduct online academic exchanges with domestic scholars and researchers. The event was held simultaneously on the online conference and live streaming platform, attracting more than 200 teachers and students to participate in.

The title of Professor Siwei Lyu’s report is “Fighting AI-synthesized fake media”. He first introduces the development status of generation technology alongside the security problems. Then he explained the effective methods to identify forged image and audio in several important aspects such as blink detection, physical clues and speech analysis, and put forward the challenge to the development of identification of deepfake algorithms from the perspective of lack of interpretability, poor robustness of data cleaning and immaturity of anti forensics technology. Finally, Professor Lyu answered several questions raised by the students and guided the audience to think about the future direction of forensics technology and the adversarial development of generation technology.

The title of Professor Ping Li’s report is “Realistic Computational Art Rendering and Synthesis”. First, he briefly introduced the current situation in the field of computational art rendering and synthesis technology, and pointed out the problems faced by traditional methods. Then, focusing on the image style transfer of structure perception, Professor Li made a specific technical introduction and achievement analysis on the problems of density enhanced perception mosaic and sketch image generation based on depth neural representation. Finally, Professor Li shared his views with the participating teachers and students about the impact of the development of deep learning on the field of art rendering and synthesis.

This activity introduced the detection and defense of synthetic media content and the latest progress of computational art rendering and synthesis, attracted the attention of many young scholars at home and abroad, and successfully concluded with the enthusiastic participation of teachers and students attending the meeting.

Facing the rapid change of forgery technology, researchers engaged in the field of forensics meet new challenges and opportunities. Based at home and embracing the world, we will actively build a broad platform in line with international standards for the exchange of scholars in the field of forensics, and strive to promote academic exchanges in cross fields at home and abroad. Finally, we sincerely thank the teachers and students who participated in this online activity. We hope you can continue to support and pay attention to our subsequent series of activities. You are also welcome to put forward suggestions and opinions on the form and content of this activity.

(Contributor: Songlin Yang, Center for Research on Intelligent Perception and Computing, Institute of Automation, Chinese Academy of Sciences)