Page 5 - Detecting deepfakes and generative AI: Report on standards for AI watermarking and multimedia authenticity workshop

List of figures and tables

Figures

Figure 1: Announcement of new standards collaboration on AI watermarking,
multimedia authenticity, and deepfake detection at the AI for Good Global
Summit 2024 ................................................................................................................................iv
Figure 2: You can help stop the spread of deepfakes ........................................................... 4
Figure 3: Robust detection of AI-synthesized human face images presented
by Touradj Ebrahimi, EPFL ......................................................................................................... 9
Figure 4: Cheaper, better, and ubiquitous deepfakes will cost governments
and enterprises billions by 2023 ............................................................................................... 9
Figure 5: Anatomy of a deepfake attack ................................................................................10
Figure 6: How to improve generalization ability of deepfake detection ..........................11
Figure 7: Content Credentials .................................................................................................13
Figure 8: Visuals generated by various AI software 6 months ago, presented
by Andrew Jenkins ....................................................................................................................13
Figure 9: Mosaic of AI-generated images shown during the workshop by
Andrew Jenks to illustrate the rapid advancement of AI in contrast with Figure
3. Warning: these images are not real. ..................................................................................14
Figure 10: Techniques for establishing media provenance and authenticity ..................15
Figure 11: Comparison between SenseTime’s solution and traditional
watermarking methods ............................................................................................................16
Figure 12: Trust elements ........................................................................................................16
Figure 13: Discussion on Standards Collaboration in Session 4........................................18

Table

Table 1: Impact of Deepfakes .................................................................................................21

iii