Dear VAFI community,

At the time of the last newsletter, the COVID-19 pandemic was raging and it is a relief that the current situation has become less dramatic even though we must remain vigilant.

As coordinator of the VICTORIA project, I am pleased to say that the consortium partners have been able to find arrangements to carry out their different tasks despite the challenging circumstances. Two public deliverables have been recently finalised and posted up on the VICTORIA Public Website. One concerns the standardisation activities carried out in the project and the results obtained to date to address the interoperability issue between the LEA tools and the variety of existing and heterogeneous solutions of the CCTV market (more here: https://www.victoria-project.eu/public-deliverables/ ). The other reports on the training process (e.g. the two live training workshops and the online platform) and assesses to what extent the end users have furthered their knowledge about video analysis, legal procedures and ethical principles (more here: https://www.victoria-project.eu/public-deliverables/ ).

It was nevertheless necessary to extend the lifetime of VICTORIA, which now has an end date of 30 November, and to turn the Final Conference into a virtual event. We are now relaunching the preparation of the VICTORIA Conference. We can now announced the new dates: 16-18 November 2020. Of course, more information on the programme and registration will be provided soon.

For now, I hope that you will take an interest in reading the selection of academic articles and technology updates provided by this third issue of the VICTORIA project newsletter. All comments and suggestions concerning the VAFI Community are always welcome so don’t hesitate to get in touch with us at VICTORIA-arttic@eurtd.com.

Wishing you all well and safe,

Luc Sonké

VICTORIA Coordinator

Scientific, academic papers and patents

04/2020

Multi-Modal Video Forensic Platform for Investigating Post-Terrorist Attack Scenarios

… The forensic investigation of a terrorist attack poses a significant challenge to the investigative authorities, as often several thousand hours of video footage must be viewed. Large scale Video Analytic Platforms (VAP) assist law enforcement agencies (LEA) in identifying suspects and securing evidence. Current platforms focus primarily on the integration of different computer vision methods and thus are restricted to a single modality. We present a video analytic platform that integrates visual and audio analytic modules and fuses information from surveillance cameras and video uploads from eyewitnesses. Videos are analyzed according their acoustic and visual content. Specifically, Audio Event Detection is applied to index the content according to attack-specific acoustic concepts. Audio similarity search is utilized to identify similar video sequences recorded from different perspectives. Visual object detection and tracking are used to index the content according to relevant concepts. Innovative user-interface concepts are introduced to harness the full potential of the heterogeneous results of the analytical modules, allowing investigators to more quickly follow-up on leads and eyewitness reports … More…

SCHINDLER, Alexander, LINDLEY, Andrew, JALALI, Anahid, et al. Multi-Modal Video Forensic Platform for Investigating Post-Terrorist Attack Scenarios. arXiv preprint arXiv:2004.01023, 2020.

04/2020

Content Security for a Video Stream

“A video capturing device configured to capture video data, the image capturing device comprising: an image processing pipeline configured to generate a video stream from the captured video data; an encryption key activating module configured to, upon request, activate an encryption key and deactivate any previously activated encryption key used for encrypting the video stream, such that only one encryption key is activated to be used for encryption of the video stream at each instance in time; an encryption scheduler module configured to, at a predetermined time interval, instruct the encryption key activating module to activate an encryption key; an event handling module configured to, as a response to receiving indications of a detected event, instruct the encryption key activating module to activate an encryption key; an encryption module configured to encrypt the video stream using the one activated encryption key” More…

FARZONE, Mazdak et KESKIKANGAS, Axel. Content security for a video stream. U.S. Patent Application No 16/552,830, 2 apr. 2020.

04/2020

Knife and Threat Detectors

“Despite rapid advances in image-based machine learning, the threat identification of a knife wielding attacker has not garnered substantial academic attention. This relative research gap appears less understandable given the high knife assault rate (>100,000 annually) and the increasing availability of public video surveillance to analyze and forensically document. We present three complementary methods for scoring automated threat identification using multiple knife image datasets, each with the goal of narrowing down possible assault intentions while minimizing misidentifying false positives and risky false negatives. To alert an observer to the knife-wielding threat, we test and deploy classification built around MobileNet in a sparse and pruned neural network with a small memory requirement (< 2.2 megabytes) and 95% test accuracy. We secondly train a detection algorithm (MaskRCNN) to segment the hand from the knife in a single image and assign probable certainty to their relative location. This segmentation accomplishes both localization with bounding boxes but also relative positions to infer overhand threats. A final model built on the PoseNet architecture assigns anatomical waypoints or skeletal features to narrow the threat characteristics and reduce misunderstood intentions. We further identify and supplement existing data gaps that might blind a deployed knife threat detector such as collecting innocuous hand and fist images as important negative training sets. When automated on commodity hardware and software solutions one original research contribution is this systematic survey of timely and readily available image-based alerts to task and prioritize crime prevention countermeasures prior to a tragic outcome” More…

NOEVER, David A. et NOEVER, Sam E. Miller. Knife and Threat Detectors. arXiv preprint arXiv:2004.03366, 2020

03/2020

Comprehensive Study of Various Techniques for Object Detection and Tracking in Video

“The methods for object detection helps in identifying the presence of an object whether in an image or a video, in the area of computer vision. They are helpful in many applications such as man-machine interfaces, traffic monitoring, transportation, manufacturing, target tracking, robotics, health care domain and various indoor-outdoor based video surveillance application. Many algorithms are used for implementing methods for object detection, classification, representation & essentially tracking. This paper has focused on various methods for object detection and tracking. The efficacy of this paper work is shown in discussion for various methods.”  More…

SURI, Ajay, SHARMA, Sudhir Kumar, et YADAV, Dileep Kumar. Comprehensive Study of Various Techniques for Object Detection and Tracking in Video. Available at SSRN 3548453, 2020.

02/2020

A Review on Big Data Analytics to Smart Cities

“…Many governments are considering adopting the smart city concept in their cities and implementing big data applications that support smart city components to achieve the required level of sustainability and progress the living standards. Smart cities exploit multiple technologies to improve the performance of health, transportation, energy, education, Safety and water services leading to higher levels of comfort of their citizens. This involves reducing costs and resource consumption in addition to more effectively and actively engaging with their citizens. One of the recent technologies that have a huge latent to enhance smart city services is big data analytics. Big data is an emerging area of research and its eventual applications in smart cities are widely recognized. As digitization has become an essential part of everyday life, data collection has resulted in the accretion of huge amounts of data that can be used in various beneficial application domains. Effective analysis and exploitation of big data is a key factor for success in many business and service domains, including the smart city domain. This paper reviews the applications of big data to support smart cities. It discusses and compares different definitions of the smart city and big data and explores the opportunities, challenges and benefits of incorporating big data applications for smart cities. In addition, it attempts to recognize the requirements that support the realization of big data applications for smart city services. ...”  More…

UTHRA, R. et BEGAM, M. Rabiyathul Bagiriya. A Review on Big Data Analytics to Smart Cities.

01/2020

Behavioral Measurements in a Video Stream Focalized on Keywords

“A system and method for analyzing behavior in a video is described. The method includes extracting a plurality of salient fragments of a video; associating a time anchor with an utterance of a first keyword in an audio track associated with the video; generating a focalized visualization, based on the time anchor, from one or more of the plurality of salient fragments of the video; tagging a human subject in the focalized visualization with a unique identifier; and analyzing behavior of the human subject, using the focalized visualization, to generate a behavior score associated with the unique identifier and the first keyword” More…

NARASIMHA, Ramya et GONZALEZ-BANOS, Hector H. Behavioral Measurements in a Video Stream Focalized on Keywords. U.S. Patent Application No 16/588,653, 23 jan. 2020.

Technology watch & Press release

May 2020 –www.policeone.com/

Knowledge is never on lockdown: learn with FARO Online

Due to the COVID-related circumstances, the public safety experts at FARO have designed a series of free virtual classrooms for solutions that can help you maximize your forensic investigations.. More…

May 2020 –www.policeone.com/

Motorola Solutions provides new features to its pronto mobile solution to help police forces practice social distancing

LONDON, UK – As a result of the urgent need to protect the safety of the public, alongside demand from UK police forces to be able to do so while respecting social distancing, Motorola Solutions is now offering new Pronto mobile applications to allow engagement with the public, whilst respecting 2 metres of distance, ... More…

May 2020 – www.securityworldmarket.com/

AI poses both challenges & opportunities for security community

“The Royal United Services Institute, or RUSI, was recently commissioned by GCHQ to conduct an independent research study into the use of artificial intelligence (AI) for national security purposes. The study revealed that AI could offer the intelligence and security communites many opportunities, but that there are also many barriers and challenges to overcome with the technology.”... More…

April 2020 –ipvm.com/

30 Million Criminal Face Database Tested (Captis Intelligence)

“30 million criminal mugshots are now available for facial recognition searches by subscribing to Captis Intelligence. We subscribed to Captis' service, connected to Milestone XProtect, to see how well it worked, including:

How accurate is the face recognition?

• What faces are included in the database?
• What face details does Captis identify?
• What criminal data does Captis include?
• What issues were found?

How does it compare to other facial recognition systems?”... More…

April 2020 – www.securityworldmarket.com/

The impact of COVID-19 on Homeland Security Management

“Post COVID-19, the size of the "COVID-19 Impact on Homeland Security Management Market by Technology and Region - Global Forecast to 2025" under realistic scenario is estimated by research company, Marketsandmarkets, to grow from USD 1,035 billion in 2020 and projected to reach USD 1,081 billion by 2025, at a CAGR of 0.9%.”... More…

March 2020 – www.securityworldmarket.com/

Industry first subject ID embedded in VMS platform

“Captis Intelligence has introduced an innovative ODSI (On-Demand Subject Identification) application designed exclusively to be embedded in Milestone’s Xprotect. VMS Captis ODSI (On-Demand Subject Identification) provides Xprotect users with a completely embedded solution to instantly identify individuals in seconds with one click. An industry first, Captis ODSI provides exclusive access to the Captis Database, the largest private database of criminals in North America with over 30 million subjects”... More…