Developing automated methods to detect and match face and voice biometrics in child sexual abuse videos

Computer screen with electronic fingerprint
Abstract

The proliferation of child sexual abuse material (CSAM) is outpacing law enforcement’s ability to address the problem. In response, investigators are increasingly integrating automated software tools into their investigations. These tools can detect or locate files containing CSAM, and extract information contained within these files to identify both victims and offenders.

Software tools using biometric systems have shown promise in this area but are limited in their utility due to a reliance on a single biometric cue (namely, the face). This research seeks to improve current investigative practices by developing a software prototype that uses both faces and voices to match victims and offenders across CSAM videos. This paper describes the development of this prototype and the results of a performance test conducted on a database of CSAM. Future directions for this research are also discussed.

References

URLs correct as at February 2022

Bourke ML & Craun SW 2014. Secondary traumatic stress among Internet Crimes Against Children Task Force personnel. Sexual Abuse: A Journal of Research and Treatment 26(6): 586–609. https://doi.org/10.1177/1079063213509411

Brewer R, Westlake B, Hart T & Arauza O 2021. The ethics of web crawling and web scraping in criminological research: Navigating issues of consent, privacy and other potential harms associated with automated data collection. In A Lavorgna & T Holt (eds), Researching cybercrimes. Cham: Palgrave: 435–456. https://doi.org/10.1007/978-3-030-74837-1_22

Brown R, Napier S & Smith RG 2020. Australians who view live streaming of child sexual abuse: An analysis of financial transactions. Trends & issues in crime and criminal justice no. 589. Canberra: Australian Institute of Criminology. https://doi.org/10.52922/ti04336

Burns CM, Morley J, Bradshaw R & Domene J 2008. The emotional impact on coping strategies employed by police teams investigating internet child exploitation. Traumatology 14(2): 20–31. https://doi.org/10.1177/1534765608319082

Bursztein E, Clarke E, DeLaune M, Elifff DM, Hsu N, Olson L, Shehan J, Thakur M, Thomas K & Bright T 2019. Rethinking the detection of child sexual abuse imagery on the internet. World Wide Web Conference, 13 May, pp 2601–2607. https://doi.org/10.1145/3308558.3313482

Canadian Centre for Child Protection 2021. Project Arachnid. https://projectarachnid.ca/en/

Chowdhury A, Atoum Y, Truan L, Liu X & Ross A 2018. MSU-AVIS dataset: Fusing face and voice modalities for biometric recognition in indoor surveillance videos. Proceedings of the 24th IAPR International Conference on Pattern Recognition (ICPR), Beijing, pp 3567–3573. https://doi.org/10.1109/ICPR.2018.8545260

Chowdhury A, Cozzo A & Ross A 2020. JukeBox: A multilingual singer recognition dataset. Proceedings of Interspeech Conference 2020, Shanghai, China. https://doi.org/10.21437/Interspeech.2020-2972

Chowdhury A & Ross A 2020. Fusing MFCC and LPC features using 1D Triplet CNN for speaker recognition in severely degraded audio signals. IEEE Transactions on Information Forensics and Security 15: 1616–1629. https://doi.org/10.1109/TIFS.2019.2941773

Chung JS, Nagrani A & Zisserman A 2018. VoxCeleb2: Deep speaker recognition. Proceedings of Interspeech Conference 2018. https://doi.org/10.21437/Interspeech.2018-1929

Council of Europe 2021. Automated detection of child sexual abuse materials. Octopus Conference 2021. https://www.coe.int/en/web/cybercrime/workshop-automated-detection-of-child-sexual-abuse-materials

Cubitt T, Napier S & Brown R 2021. Predicting prolific live streaming of child sexual abuse. Trends & issues in crime and criminal justice no. 634. Canberra: Australian Institute of Criminology. https://doi.org/10.52922/ti78320

Dance GJX & Keller MH 2020. Tech companies detect a surge in online videos of child sexual abuse. New York Times, 20 February. https://www.nytimes.com/2020/02/07/us/online-child-sexual-abuse.html

de Castro Polastro M & da Silva Eleuterio PM 2010. NuDetective: A forensic tool to help combat child pornography through automatic nudity detection. 2010 Workshops on Database and Expert Systems Applications, pp 349–353. https://doi.org/10.1109/DEXA.2010.74

Foley J, Louise K & Massey D 2020. The ‘cost’ of caring in policing: From burnout to PTSD in police officers in England and Wales. The Police Journal: Theory, Practice, and Principles 94(3): 298–315. https://doi.org/10.1177/0032258X20917442

Gangwar A, González-Castro V, Alegre E & Fidalgo E 2021. AttM-CNN: Attention and metric learning based CNN for pornography, age and child sexual abuse (CSA) detection in images. Neurocomputing 445: 81–104. https://doi.org/10.1016/j.neucom.2021.02.056

Hernandez-Ortega J, Galbally J, Fierrez J, Haraksim R & Beslay L 2019. FaceQnet: Quality assessment for face recognition based on deep learning. Proceedings of the 12th International Conference on Biometrics, Crete. https://doi.org/10.1109/ICB45273.2019.8987255

Huang GB, Mattar M, Berg T & Learned-Miller E 2008. Labeled Faces in the Wild: A database for studying face recognition in unconstrained environments. Workshop on Faces in ‘Real-Life’ Images: Detection, Alignment, and Recognition. https://hal.inria.fr/inria-00321923

Internet Watch Foundation 2021. The annual report 2020. https://annualreport2020.iwf.org.uk/

Interpol 2022. Victim identification. https://www.interpol.int/en/Crimes/Crimes-against-children/Victim-identification

Interpol 2018. Towards a global indicator on unidentified victims in child sexual exploitation material: Technical report. https://ecpat.org/resource/technical-report-towards-a-global-indicator-on-unidentified-victims-in-child-sexual-exploitation-material/

Islam M, Watters PA & Yearwood J 2011. Real-time detection of children’s skin on social networking sites using markov random field modelling. Information Security Technical Report 16(2): 51–81. https://doi.org/10.1016/j.istr.2011.09.004

Islam M, Watters P, Mahmood AN & Alazab M 2019. Toward detection of child exploitation material: A forensic approach. In M Alazab & M Tang (eds), Deep learning applications for cyber security. Cham: Springer: 221–246. https://doi.org/10.1007/978-3-030-13057-2_11

Jain AK, Klare B & Ross A 2015. Guidelines for best practices in biometric research. Proceedings of the 8th IAPR International Conference on Biometrics (ICB), Phuket, Thailand, pp 541–545. https://doi.org/10.1109/ICB.2015.7139116

Macedo J, Costa F & dos Santos JA 2018. A benchmark methodology for child pornography detection. 2018 31st SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI), Paraná, Brazil, pp 455–462. https://doi.org/10.1109/SIBGRAPI.2018.00065

Maxim D, Orlando S, Skinner K & Broadhurst R 2016. Online child exploitation material: Trends and emerging issues. Canberra: Australian National University Cybercrime Observatory and Office of the Children’s eSafety Commissioner. https://doi.org/10.2139/ssrn.2861644

Monk B, Allsup R & Frank R 2015. LECENing places to hide: Geo-mapping child exploitation material. 2015 IEEE International Conference on Intelligence and Security Informatics (ISI), Baltimore, pp 73–78. https://doi.org/10.1109/ISI.2015.7165942

Moser A, Rybnicek M & Haslinger D 2015. Challenges and limitations concerning automatic child pornography classification. Proceedings of the 10th International Conference on Computer Vision Theory and Applications, Berlin, pp 492–497. https://doi.org/10.5220/0005344904920497

National Center for Missing & Exploited Children 2020. CyberTipline. https://www.missingkids.org/gethelpnow/cybertipline

Park TJ, Kanda N, Dimitriadis D, Han KJ, Watanabe S & Narayanan S 2022. A review of speaker diarization: Recent advances with deep learning. Computer Speech & Language 72. https://doi.org/10.1016/j.csl.2021.101317

Phippen A & Bond E 2020. Image recognition in child sexual exploitation material: Capabilities, ethics and rights. In H Jahankhani, B Akhgar, P Cochrane & M Dastbaz (eds), Policing in the era of AI and smart societies. Advanced Sciences and Technologies for Security Applications. Cham: Springer: 179–198. https://doi.org/10.1007/978-3-030-50613-1_8

Powell M, Cassematis P, Benson M, Smallbone S & Wortley R 2015. Police officers’ perceptions of their reactions to viewing internet child exploitation material. Journal of Police and Criminal Psychology 30(2): 103–111. https://doi.org/10.1007/s11896-014-9148-z

Ross A, Nandakumar K & Jain AK 2006. Handbook of multibiometrics vol. 6. Springer. https://doi.org/10.1007/0-387-33123-9

Sae-Bae N, Sun X, Sencar HT & Memon ND 2014. Towards automatic detection of child pornography. 2014 IEEE International Conference on Image Processing, Paris, pp 5332–5336. https://doi.org/10.1109/ICIP.2014.7026079

Sanchez L, Grajeda C, Baggili I & Hall C 2019. A practitioner survey exploring the value of forensic tools, AI, filtering, & safer presentation for investigating child sexual abuse material (CSAM). Digital Investigation 29: S124–S142. https://doi.org/10.1016/j.diin.2019.04.005

Seigfried-Spellar KC 2018. Assessing the psychological well-being and coping mechanisms of law enforcement investigators vs. digital forensic examiners of child pornography investigations. Journal of Police and Criminal Psychology 33(3): 215–226. https://doi.org/10.1007/s11896-017-9248-7

Srinivas N, Ricanek K, Michalski D, Bolme DS & King M 2019. Face recognition algorithm bias: Performance differences on images of children and adults. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, pp 2269–2277. https://doi.org/10.1109/CVPRW.2019.00280

Tanwar S, Sudhanshu T, Kumar N & Obaidat MS 2019. Online signature-based biometric recognition. In MS Obaidat, I Traore I & I Woungang (eds), Biometric-based physical and cybersecurity systems. Cham: Springer: 535–570. https://doi.org/10.1007/978-3-319-98734-7_10

Ulges A & Stahl A 2011. Automatic detection of child pornography using color visual words. 2011 IEEE International Conference on Multimedia and Expo, Barcelona, pp 1–6. https://doi.org/10.1109/ICME.2011.6011977

Vitorino P, Avila S, Perez M & Rocha A 2018. Leveraging deep neural networks to fight child pornography in the age of social media. Journal of Visual Communication and Image Representation 50: 303–313. https://doi.org/10.1016/j.jvcir.2017.12.005

WebRTC 2017. Real-time communication for the web. https://webrtc.org/

Westlake BG, Bouchard M & Frank R 2017. Assessing the validity of automated webcrawlers as data collection tools to investigate online child sexual exploitation. Sexual Abuse: A Journal of Research and Treatment 29(7): 685–708. https://doi.org/10.1177/1079063215616818

Yaqub W, Mohanty M & Memon N 2018. Encrypted domain skin tone detection for pornographic image filtering. 2018 15th IEEE International Conference on Advanced Video and Signal Based Surveillance, Auckland, pp 1–5. https://doi.org/10.1109/AVSS.2018.8639350

Yiallourou E, Demetriou R & Lanitis A 2017. On the detection of images containing child-pornographic material. 24th International Conference on Telecommunications, Cyprus. https://doi.org/10.1109/ICT.2017.7998260

Yiu SY, Malec C & Michalski D 2021. Performance of facial recognition algorithms for the 5RD combating child exploitation network. DSTG-CR-2021-0160. Edinburgh, Australia: Defence Science and Technology Group