Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
110 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

GenAI Mirage: The Impostor Bias and the Deepfake Detection Challenge in the Era of Artificial Illusions (2312.16220v2)

Published 24 Dec 2023 in cs.CV

Abstract: This paper examines the impact of cognitive biases on decision-making in forensics and digital forensics, exploring biases such as confirmation bias, anchoring bias, and hindsight bias. It assesses existing methods to mitigate biases and improve decision-making, introducing the novel "Impostor Bias", which arises as a systematic tendency to question the authenticity of multimedia content, such as audio, images, and videos, often assuming they are generated by AI tools. This bias goes beyond evaluators' knowledge levels, as it can lead to erroneous judgments and false accusations, undermining the reliability and credibility of forensic evidence. Impostor Bias stems from an a priori assumption rather than an objective content assessment, and its impact is expected to grow with the increasing realism of AI-generated multimedia products. The paper discusses the potential causes and consequences of Impostor Bias, suggesting strategies for prevention and counteraction. By addressing these topics, this paper aims to provide valuable insights, enhance the objectivity and validity of forensic investigations, and offer recommendations for future research and practical applications to ensure the integrity and reliability of forensic practices.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (94)
  1. Cognitive biases and decision-making strategies in times of change: A systematic literature review. Management Decision 59, 638–652.
  2. Reverse engineering of generative models: Inferring model hyperparameters from generated images. IEEE Transactions on Pattern Analysis and Machine Intelligence .
  3. Multimedia forensics: Discovering the history of multimedia contents, in: Proceedings of the 17th International Conference on Computer Systems and Technologies 2016, pp. 5–16.
  4. The impact of cognitive biases on professionals’ decision-making: A review of four occupational areas. Frontiers in psychology 12, 802439.
  5. Is forensic evidence impartial? cognitive biases in forensic analysis. Criminal Psychology and the Criminal Justice System in India and Beyond , 215–227.
  6. How Russia is losing — and winning — the information war in Ukraine. NPR Accessed: 2023-12-13.
  7. A risk-based approach to cognitive bias in forensic science. Science & Justice : Journal of the Forensic Science Society 59 5, 533–543. doi:10.1016/J.SCIJUS.2019.04.003.
  8. Image-to-Image Translation via Group-Wise Deep Whitening-and-Coloring Transformation, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 10639–10647.
  9. StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-image Translation, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 8789–8797.
  10. Social Media Warfare Is Being Invented in Ukraine. https://www.cigionline.org/articles/social-media- warfare-is-being-invented-in-ukraine/. Accessed: 2023-12-13.
  11. On the generalization of deep learning models in video deepfake detection. Journal of Imaging 9, 89.
  12. Combining efficientnet and vision transformers for video deepfake detection, in: International Conference on Image Analysis and Processing, Springer. pp. 219–229.
  13. Cognitive bias research in forensic science: A systematic review. Forensic Science International 297, 35–46.
  14. On the detection of synthetic images generated by diffusion models, in: ICASSP 2023-2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), IEEE. pp. 1–5.
  15. Pattern recognition after image processing of low-contrast images, the case of the shroud of turin. Pattern Recognition 46, 1964–1970.
  16. Implicit identity leakage: The stumbling block to improving deepfake detection generalization, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 3994–4004.
  17. Case information biases evaluations of video-recorded eyewitness identification evidence. Journal of Applied Research in Memory and Cognition .
  18. Practical solutions to cognitive and human factor challenges in forensic science. Forensic Science Policy & Management: An International Journal 4, 105 – 113. doi:10.1080/19409044.2014.901437.
  19. Cognitive bias in forensic pathology decisions. Journal of Forensic Sciences 66, 1751–1757.
  20. Meta-analytically quantifying the reliability and biasability of forensic experts. Journal of Forensic Sciences 53, 900–903.
  21. Cognitive and human factors in expert decision making: Six fallacies and the eight sources of bias. Analytical Chemistry 92, 7998–8004.
  22. Linear sequential unmasking–expanded (lsu-e): A general approach for improving decision making as well as minimizing noise and bias. Forensic Science International: Synergy 3, 100161.
  23. Watch your up-convolution: Cnn based generative deep neural networks are failing to reproduce spectral distributions, in: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 7887–7896. doi:10.1109/CVPR42600.2020.00791.
  24. Fourier spectrum discrepancies in deep network generated images, in: Larochelle, H., Ranzato, M., Hadsell, R., Balcan, M.F., Lin, H. (Eds.), Advances in Neural Information Processing Systems, Curran Associates, Inc.. pp. 3022–3032.
  25. Art and the science of generative ai: A deeper dive. arXiv preprint arXiv:2306.04141 .
  26. An innovative tool for uploading/scraping large image datasets on social networks. arXiv e-prints , arXiv–2311.
  27. The double superficiality of the frontal image of the turin shroud. Journal of Optics A: Pure and Applied Optics 6, 491.
  28. Cognitive bias: Steering conclusions irrationally. https://blog.ampedsoftware.com/2021/04/20/cognitive-bias-steering-conclusions-irrationally. Accessed: 2023-11-07.
  29. Forensic Science Regulator, 2020. Cognitive bias effects relevant to forensic science examinations. https://www.gov.uk/government/publications/cognitive-bias-effects-relevant-to-forensic-science-examinations. Accessed: 2023-11-07.
  30. Do evidence submission forms expose latent print examiners to task-irrelevant information? Forensic Science International 297, 236–242.
  31. Towards discovery and attribution of open-world gan generated images, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 14094–14103.
  32. Fighting deepfakes by detecting gan dct anomalies. Journal of Imaging 7, 128. doi:10.3390/jimaging7080128.
  33. A classification engine for image ballistics of social data, in: Image Analysis and Processing-ICIAP 2017: 19th International Conference, Catania, Italy, September 11-15, 2017, Proceedings, Part II 19, Springer. pp. 625–636.
  34. Generative adversarial nets, in: Advances in Neural Information Processing Systems, pp. 2672–2680.
  35. Overconfidence and financial decision-making: A meta-analysis. Review of Behavioral Finance doi:10.1108/rbf-01-2020-0020.
  36. Using cognitive bias modification to deflate responsibility in compulsive checkers. Cognitive Therapy and Research 38, 505–517.
  37. Applying artificial intelligence for age estimation in digital forensic investigations. arXiv:2201.03045.
  38. Deepfake detection by analyzing convolutional traces, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, pp. 666–667.
  39. Fighting deepfake by exposing the convolutional traces on images. IEEE Access 8, 165085–165098.
  40. Level up the deepfake detection: A method to effectively discriminate images generated by gan architectures and diffusion models. arXiv preprint arXiv:2303.00608 .
  41. Preliminary forensics analysis of deepfake images, in: 2020 AEIT International Annual Conference (AEIT), IEEE. pp. 1–6. doi:10.23919/AEIT50178.2020.9241108.
  42. On the exploitation of deepfake model recognition, in: 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), pp. 61–70. doi:10.1109/CVPRW56347.2022.00016.
  43. Consumers’ overconfidence biases in relation to social exclusion. The Journal of Asian finance, economics, and business 7, 303–308.
  44. Deep residual learning for image recognition, in: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 770–778.
  45. Deepfake detection algorithm based on improved vision transformer. Applied Intelligence 53, 7512–7527.
  46. The Development of Case Assessment and Interpretation (CAI) in Forensic Science. Ph.D. thesis. University of Abertay Dundee.
  47. Strengthening forensic dna decision making through a better understanding of the influence of cognitive bias. Science & Justice : Journal of the Forensic Science Society 57 6, 415–420. doi:10.1016/j.scijus.2017.07.005.
  48. Diverging deep learning cognitive computing techniques into cyber forensics. Forensic Science International: Synergy 1, 61–67.
  49. The forensic confirmation bias: Problems, perspectives, and proposed solutions. Journal of applied research in memory and cognition 2, 42–52.
  50. Data integrity in global clinical trials: Discussions from joint us fda and mhra uk good clinical practice workshop. Clinical Pharmacology and Therapeutics doi:10.1002/cpt.1794.
  51. Retention and transfer of cognitive bias mitigation interventions: A systematic literature study. Frontiers in Psychology 12, 629354.
  52. Cognitive bias and blindness: A global survey of forensic science examiners. Journal of Applied Research in Memory and Cognition 6, 452–459. doi:10.1016/J.JARMAC.2017.09.001.
  53. Tar: Generalized forensic framework to detect deepfakes using weakly supervised learning, in: IFIP International Conference on ICT Systems Security and Privacy Protection, Springer. pp. 351–366.
  54. Not with my name! inferring artists’ names of input strings employed by diffusion models, in: International Conference on Image Analysis and Processing, Springer. pp. 364–375.
  55. Modifying cognitive errors promotes cognitive well being: A new approach to bias modification. Journal of Behavior Therapy and Experimental Psychiatry 42, 298–308.
  56. Deepfakes in warfare: new concerns emerge from their use around the Russian invasion of Ukraine. http://theconversation.com/deepfakes-in-warfare-new-concerns-emerge-from-their-use-around-the-russian-invasion-of-ukraine-216393. Accessed: 2023-12-13.
  57. Metric learning from relative comparisons by minimizing squared residual, in: 2012 IEEE 12th International Conference on Data Mining, IEEE. pp. 978–983.
  58. Do gans leave artificial fingerprints?, in: 2019 IEEE conference on multimedia information processing and retrieval (MIPR), IEEE. pp. 506–511.
  59. Deepfakes generation and detection: State-of-the-art, open challenges, countermeasures, and way forward. Applied Intelligence 53, 3974–4026.
  60. Cognitive biases in criminal case evaluation: A review of the research. Journal of Police and Criminal Psychology 37, 101–122.
  61. The expectation-maximization algorithm. IEEE Signal Processing Magazine 13, 47–60.
  62. Cognitive bias in forensic anthropology: Visual assessment of skeletal remains is susceptible to confirmation bias. Science & Justice 54, 208–214.
  63. The cognitive underpinnings of bias in forensic mental health evaluations. Psychology, Public Policy and Law 20, 200–211. doi:10.1037/A0035824.
  64. A general model of cognitive bias in human judgment and systematic review specific to forensic mental health. Law and Human Behavior doi:10.1037/lhb0000482.
  65. Response bias (response style). The Encyclopedia of Cross-Cultural Psychology , 1098–1103.
  66. Glide: Towards photorealistic image generation and editing with text-guided diffusion models, in: International Conference on Machine Learning, PMLR. pp. 16784–16804.
  67. A road map for digital forensic research, in: First digital forensic research workshop, utica, new york, pp. 27–30.
  68. What is the function of confirmation bias? Erkenntnis 87, 1351–1376. doi:10.1007/s10670-020-00252-1.
  69. An overview on image forensics. International Scholarly Research Notices 2013.
  70. Applications of generative ai to media. SMPTE Motion Imaging Journal 132, 53–57. doi:10.5594/JMI.2023.3297238.
  71. Hierarchical Text-Conditional Image Generation with Clip Latents. arXiv preprint arXiv:2204.06125 .
  72. Ace-v and the scientific method. Journal of Forensic Identification 60, 87.
  73. Efficacy of psychological interventions targeting cognitive biases in schizophrenia: A systematic review and meta-analysis. Clinical Psychology Review 78, 101854. doi:https://doi.org/10.1016/j.cpr.2020.101854.
  74. De-fake: Detection and attribution of fake images generated by text-to-image generation models, in: Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security, pp. 3418–3432.
  75. Deep unsupervised learning using nonequilibrium thermodynamics, in: International Conference on Machine Learning, PMLR. pp. 2256–2265.
  76. A biased opinion: Demonstration of cognitive bias on a fingerprint matching task through knowledge of dna test results. Forensic Science International 276, 93–106.
  77. Bias among forensic document examiners: Still a need for procedural changes. Australian Journal of Forensic Sciences 46, 91 – 97. doi:10.1080/00450618.2013.797026.
  78. Is Russia’s Invasion Of Ukraine The First Social Media War? https://www.forbes.com/sites/petersuciu/2022/03/01/is-russias-invasion-of-ukraine-the-first-social-media-war/. Accessed: 2023-12-13.
  79. Cognitive and human factors in digital forensics: Problems, challenges, and the way forward. Digital Investigation 29, 101–108.
  80. Lay understanding of forensic statistics: Evaluation of random match probabilities, likelihood ratios, and verbal equivalents. Law and human behavior 39 4, 332–49. doi:10.1037/lhb0000134.
  81. Media forensics and deepfakes: an overview. IEEE Journal of Selected Topics in Signal Processing 14, 910–932.
  82. Face pareidolia and its neural mechanism. Advances in Psychological Science 26, 1952.
  83. M2tr: Multi-modal multi-scale transformers for deepfake detection, in: Proceedings of the 2022 International Conference on Multimedia Retrieval, pp. 615–623.
  84. Fakespotter: A simple yet robust baseline for spotting ai-synthesized fake faces, in: Proceedings of the Twenty-Ninth International Conference on International Joint Conferences on Artificial Intelligence, pp. 3444–3451.
  85. Cnn-generated images are surprisingly easy to spot… for now, in: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 8695–8704.
  86. Jumping to conclusions and delusions: The impact of discussion of the bias on the bias. Schizophrenia Research 150, 575–579. doi:10.1016/j.schres.2013.09.003.
  87. Deepfake video detection using convolutional vision transformer. CoRR abs/2102.11126. URL: https://arxiv.org/abs/2102.11126, arXiv:2102.11126.
  88. Overconfidence in judgements: the evidence, the implications and the limitations. The Journal of Prediction Markets 2, 73–90. doi:10.5750/JPM.V2I1.436.
  89. Attributing fake images to gans: Learning and analyzing gan fingerprints, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 7556–7566.
  90. Artificial fingerprinting for generative models: Rooting deepfake attribution in training data, in: Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 14448–14457.
  91. Responsible disclosure of generative models using scalable fingerprinting. arXiv preprint arXiv:2012.08726 .
  92. Detecting and simulating artifacts in gan fake images, in: 2019 IEEE International Workshop on Information Forensics and Security (WIFS), IEEE. pp. 1–6.
  93. Do you see the “face”? individual differences in face pareidolia. Journal of Pacific Rim Psychology 14. doi:10.1017/prp.2019.27.
  94. Do you see the “face”? individual differences in face pareidolia. Journal of Pacific Rim Psychology 14, e2.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (4)
  1. Mirko Casu (3 papers)
  2. Luca Guarnera (20 papers)
  3. Pasquale Caponnetto (1 paper)
  4. Sebastiano Battiato (45 papers)
Citations (6)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now