Image and video detection/classification

Interpol, Inspectoratul General AL Politiei, ZiuZ Visual Intelligence, NCIS Norway

Project CPORT is funded by the European Union. It is a 2-year project that started in Jan 2023. The CPORT portal allows law enforcement agencies to access ICCAM – a database used by national hotlines to process and exchange reports of child sexual abuse material (CSAM) submitted by the public worldwide. The ICCAM platform is hosted by INTERPOL and is an important channel into INTERPOL’s international child sexual exploitation database and list of known domains disseminating severe abuse material (IWOL).

ZiuZ Forensic, Web-IQ, Timelex, DFKI, INHOPE

The AviaTor project is funded by the European Union and is currently in its final phase. The project started in 2019 and will end in 2024. The project aimed to build a prioritisation tool / database for law enforcement processing NCMEC reports (also known as industry reports or Cyber tipline referrals). The AviaTor database provides law enforcement with the tooling to prioritise these reports. AviaTor stands for Augmented Visual Intelligence and Targeted Online Research – meaning that the AviaTor database has the functionality to use visual intelligence as well as OSINT and hash matching to de-duplicate and prioritise reports. AviaTor is currently used by 19 national law enforcement agencies.

Cybertip Canada

Rather than waiting for Project Arachnid to detect material and send a notice, industry can use Shield by Project Arachnid to quickly detect known CSAM on their service, which will, in turn, speed up its removal. Industry members that do not wish to interact directly with Shield by Project Arachnid can register their service/domain with the Canadian Centre to have any notices sent directly to them instead of being sent to their hosting provider. Other industries, such as filtering providers, can download real-time lists of URLs that are believed to currently contain CSAM for filtering purposes.

Cybertip Canada

Project Arachnid is a platform for reducing the availability of child sexual abuse material (CSAM) which includes crawling the internet for known CSAM using digital fingerprints as well as providing an API for industry members to improve upon and accelerate the detection of this harmful material on their own networks.

Dartmouth College

eGlyph is a robust-hashing algorithm that can identify previously tagged images,video, and audio. eGlyph extends photoDNA which operates only on images.

EOKM and Web-IQ

HashCheckServer is a tool for hosting providers to prevent the uploading of known child sexual exploitation material. To prevent the uploading of known child sexual exploitation material, hosting parties needed to be part of the solution as they provide one of the gateways to publishing CSAM. Based on this knowledge, EOKM took the initiative to help hosting parties to prevent the uploading of known CSAM and to identify existing material on their servers so that it can be deleted. Following this initiative, EOKM asked Web-IQ to develop software that allows hosting parties to check whether an image appears in the police’s database of known CSAM.

Facebook

Camera Vision is a machine learning tool that identifies images that contain both nudity and a child.

Google

The Content Safety API sorts through many images and prioritizes the most likely child sexual abuse material (CSAM) content for review. The classifier can target content that has not been previously confined as CSAM.

Google/Youtube

CSAI Match is used to surface potential child sexual abuse images (CSAI) for YouTube uploads as well as partner submissions.

Krunam

Krunam provides breakthrough technology that identifies and classifies previously unknown CSAM images and video at scale. Our CSAM classifier protects your online community, your employees, your brand’s reputation, and, last but not least, millions of children around the world.

Microsoft

PhotoDNA for Video builds on PhotoDNA technology, bringing all the benefits of PhotoDNA to the video environment.

Microsoft

PhotoDNA Cloud Service enables businesses to protect their assets, interests, and customers by autornaticuliy detecting and reporting the distribution of child sexual exploitation and abuse images.

NetClean

NetClean ProTective is a technical solution used to protect business mobile devices by blocking access to URLs known to contain child sexual abuse material (CSAM). The uniquely combined and continually updated URL list, from some of the world’s primary sources, makes it an effective protective solution.

Qumodo

Vigil At automatically detects, evaluates and categorizes child sexual abuse imagery. Thu system is capable of determining the severity of the sexual act in the image (Using the legacy UK 1-5 Category SAP Scale or the current UK 3 Categories A-C). The tool is available as part of Qumodo Classify, via a Cloud API or via a stand alone API for tools such as Griffeye Analyze. The tool scales linearly and can categorize millions of images per hour

Qumodo

Qumodo Classify is a software product which dramatically speeds up image categorization for content moderation for law enforcement purposes white reducing the psychological impact on the users, saving both time and money. This tool has been specialty designed around the latest AI technology, as well as 21 core psychology principals to form an effective Human/ AI team. As well as reducing the psychological impact on the user, the tool is designad to prevent human bias from influencing the system and creates the ability for the software to learn new abilities over time, further reducing the time needed to manually classify large amounts of data. When paired with the Vigil AI Classifier (vigil.ai) it becomes a powerful tool in the fight against child sexual exploitation and abuse (CSEA) content.

Thorn

The programme is developing artificial intelligence classifiers, which are working on automating the detection of CSAM, creating a global standard for labeling, connecting, and organizing the world's data to help identify victims faster.

Thorn

Safer is a child sexual abuse material (CSAM) detection, review, removal and reporting pipeline. It allows small and medium sized companies to have the same CSAM fighting tools as the largest ones.

Web-IQ

WebIQ has developed a serverless CSA detection AI, that scans websites and hashes images, and compares the hash to know CSA images. If there is a match, an automatic Notice of Takedown is sent to the host of the server of the website.

Two hat

Cease.AI for Law Enforcement helps reduce investigator workloads and reduce mental stress by sorting, flagging and removing non-CSAM, allowing investigators to focus their efforts on new child abuse images. investigators upload case images, run their hash lists to eliminate known material, then let the AI identify, suggest a label, and prioritize images that contain previously uncatalogued CSAM.

Videntifier Technologies

Large-scale video identification, which allows searching databases of videos. Videntifier TM Forensic, radically improves the forensic video identification process, by providing law enforcement agencies with a robust, fast and easy-to-use video identification system. Using this ser-vice, a single mouse-click is sufficient to automatically scan an entire storage device and classify all videos.

Qumodo

Qumodo Discover is a search engine for evidence. This technology uses AI to intelligently find connections within digital data at a web scale where users can query the system with an image, and it will find other images containing the same places, people, faces and objects. It has been tested with more than 100 million images and validated in a victim ID environment, with child sexual exploitation and abuse (CSEA). content

Semantics 21

S21 Laseri-v is a specialist tool for the mass categorization and analysis of child sexual abuse (CSA) videos. LASERi-v (Videos) contains proprietary workflows and algorithms to perform victim identification and investigation. The application is fully customisable, Project VIC premier certified and used by Police forces across the world for the investigation of child sexual abuse and other major crime types. (S2i LASERi-v (Videos) is available as part of the S2i LASERi suite).

Semantics 21

The S2i Laseri Suite is a specialist tool for the mass categorization and analysis of child sexual abuse imagery (CSAI) and animation. LASERi-p Pictures) contains proprietary workflows and algorithms to perform victim identification and quality investigations. The application is fully customisable, Project VIC premier certified and used by police forces across the world for the investigation of child sexual abuse and other major crime types. (S21 LASERi-p (Pictures) is available as part of the S21 LASERi suite).

Griffeye

Analyze DI allows users to import, process and review complex information as well as pre-categorise known data and correlate metadata with open sources on the internet.

NetClean

NetClein ProActive works similarly to an anti-virus program but instead of detecting a virus, ProActive detects images and videos that law enforcement agencies have classified as child sexual abuse material.

Project VIC

Project VIC is a community of investigators, victim identification specialists, application developers, scientists and strategic partners dedicated to developing and levering the newest technologies and victim-centric methods of investigation to improve success rate in rescuing victims of child sexual exploitation and trafficking. Proiect VIC has co-created an ecosystem of shared and standardized technologies specifically chosen to combat crimes involving massive amounts of images and video. Data from these respective technologies is processed using the Project VAC Cloud and redistributed back to the greater community working on crimes against children. The following tools are part of the Project VIC ecosystem: Project VIC Hash Cloud Provided by Project VIC International and Hubstream Inc. hosts over 65 million CSAM and non CSAM hashes- Access encryption Provided by Project VIC. PhotoDNA for Images by Microsoft is made available to law enforcement through a sub-license agreement with Project VIC LLC PhotoDMA for Video by Microsoft is made available to law enforcement through a sub-license agreement with Project VIC LLC Ft Video Fingerprinting by Friend MTB is made available to law enforcement through a sub-license agreement to Project VIC LLC. VICS-Odata Model Licensing is made available to law enforcement through a sub-license agreement with Project VIC LLC international Image Classifies Standard- (IC-3) Machine Learning Modelling is made available through Project VIC International Global Alert System- ViC5 Exif Matching across global platforms to identify similarities in imagery for Victim Identification. License included with ViC5

Interpol

Our International Child Sexual Exploitation (ICSE) image and video database is an intelligence and investigative tool, which allows specialized investigators to share data on cases of child sexual abuse. Using image and video comparison software, investigators are instantly able to make connections between victims, abusers and places. The database avoids duplication of effort and saves precious time by letting investigators know whether a series of images has already been discovered or identified in another country, or whether it has similar features to other images. It also allows specialized investigators from more than 64 countries to exchange information and share data with their colleagues across the world. By analysing the digital, visual and audio content of photographs and videos, victim identification experts can retrieve clues, identify any overlap in cases and combine their efforts to locate victims of child sexual abuse.

Global Emancipation Network

Global Emancipation Network grants Minerva access to organizations with valid counter-trafficking missions in order to host and explore millions of trafficking-related data records and use customized search, alerting, geo-location and other platform capabilities. Using Splunk Enterprise and third-party integrated technologies, Minerva protects case-sensitive information and monitors usage patterns to help keep user information private, safe and secure. Leading public, private and nonprofit organizations have already been accepted to the platform as early users. Minerva is equipped with data-processing capabilities to extract and organize information from a variety of data sources. Capabilities include: Advertisement analysis: Analyzes advertisements from the deep and open web, where most trafficking cases originate, and extracts data such as user, location, account and other identifying information. Image processing tools: Processes images of victims to reduce the time users spend analyzing photographs and manually linking them to advertisements. Minerva integrates image analysis tools to tag photographs with characteristics to expedite database search, and reverse image search to identify similar images. Text analysis tools and natural language processing: Extracts text in images from advertisements and flags correlations with missing persons reports and other valuable information. Multi-tenant system informing trend analysis: Allows all Minerva users to store their information securely on the same database at the same time, enabling secure, multi-agency collaboration on shared investigations.

BlueBear

LACE enables investigators to extract more visual data and efficiently categorize vast amounts of image and video data from confiscated devices using zero false positive visual matching, dramatically reducing (typical 90% case reduction) the time and resources required to bring a case to court. LACE is compatible with international standards for sharing metadata and hash signatures like ICSE, ProjectVic md UK-CAID. LACE Standalone Carver is software used for carving pictures and videos from any digital forensically sound manner. The LACE FORENSIC CARVER recognizes over 180 file types covering image, video, text documents, archive, and database files. Using this light tool, forensic experts will be able to extract evidence and import it using the JSON standard in to all the various tools they choose to use to analyse and review the evidence.

Semantics 21

S21 Laseri-t is a specialist tool for onsite (live box) and bootable (dead box) triage of laptops and computers able to detect and analyse imagery and videos. LASERi-t (Triage) contains prophet workflows and algorithms for intelligence driven search capabilities, compatibility with a wide range of intelligence sources , customisable reporting and both simple and advanced user interface modes for all tiers of users. The application is fully customisable, Project VIC premier certified and used by Police forces across the world for the investigation of child sexual abuse and other major crime types. (S21 LASERi-t Triage) is available as part of the S21 LASERi suite).