One of the most important threats to today’s civilization is terrorism, which has affected the quality of lives of people in the whole world. Terrorism is defined as the use of intentional indiscriminate and illegal power and violence for creating terror amongst the general population in order to gain some political, monetary, religious, or legal objectives.
Terrorist attacks typically involve high lethality and destructive power and directly cause massive casualties and property losses. In addition, they bring tremendous psychological pressure on people. In summary, terrorist attacks result in social unrest to a certain extent, obstructing the regular order of work and life and thus greatly hindering economic development. In the last 20 years, terrorism has led to hundreds of thousands of deaths and massive economic, political, and humanitarian crises in several regions of the world. Terrorism, in fact, undermines states’ stability, peace, and cooperation between countries, in addition to economic development and basic human rights.
Antiterrorism is an important part of global security governance, which is a sustainability issue that guarantees global security development. The multifaceted nature of terrorism, characterized by a myriad of ideologies, motives, actors, and objectives, poses a challenge to governments, institutions, and policy-makers around the world. The U.S. government spends half a trillion dollars annually to research and combat terrorism. From 2000 to 2015, 61 new terror groups emerged each year, on average, leading to an 800% increase in global terror attacks, according to data from GTD.
Given its salience and relevance, the United Nations includes the prevention of terrorism (along with violence and crime) as a target of the sixteenth Sustainable Development Goal, which specifically frames the promotion of peaceful and inclusive societies. Most states focus on preventing terrorist attacks, rather than reacting to them. As such, prediction is already central to effective counterterrorism.
There are two means to prevent terrorist attacks. One is deterrence: through the protection of infrastructure, the application of security checks and the promise of punishment. Another is the denial of the ability to conduct attacks: by apprehending terrorists before their plots come to fruition, countering recruitment and radicalization of future terrorists, and placing restrictions on the movement and freedom of individuals.
The analysis and prediction of terrorist attacks support targeted attacks on terrorist groups and provide valuable information for antiterrorism and terrorism prevention operations, enabling authorities to find new or hidden terrorists as soon as possible to reduce human and property losses, prevent problems and improve the security and stability of social life.
Police agencies are using facial and object recognition technology for counterterrorism operations. Video footage played a key role in finding the culprits responsible for the November 2015 Paris attacks, with a CCTV video at Brussels airport used to pin down one suspect. DoD and Intelligence agencies collect loads of data from satellites, drones and Internet-of-things devices. But it needs help making sense of the intelligence and analyzing it quickly enough so it can be used in combat operations. But, the sheer volume of video content produced makes identifying, assembling, and delivering actionable intelligence — from multiple sources and across thousands of hours of footage — a habitually long, laborious process.
According to the Engineering and Physical Science and Research Council: Artificial Intelligence technologies aim to reproduce or surpass abilities (in computational systems) that would require ‘intelligence’ if humans were to perform them. These include: learning and adaptation; sensory understanding and interaction; reasoning and planning; optimisation of procedures and parameters; autonomy; creativity; and extracting knowledge and predictions from large, diverse digital data.
Now defense and intelligence agencies are leveraging artificial intelligence (AI) and machine learning to automatically identify video objects of interest. They need powerful artificial intelligence software tools that the tech industry is advancing at a fast pace. The U.S. military has already spent $7.4 billion on AI to streamline and speed up video analysis in the conflict against ISIS.
AI in Video Analytics
The vast amount of digital information now generated by the average individual means that more of this routine activity could be understood through analysis. Sources include communications metadata and internet connection records, but also extend to location and activity tracking, purchases and social media activity.
Automated data analytics are used to support the activities of the intelligence and security services, particularly through data visualization. Algorithms prioritize terrorist suspects, and routinely assess the risk of air-travel passengers. Information can be collected and stored by default, to be analysed at a later time with a view to revealing patterns and links that expose terrorist networks or suspicious activities. Machine learning approaches allow the interpretation and analysis of otherwise inaccessible patterns in large amounts of data. These approaches may involve filtering, analysis of relationships between entities, or more sophisticated image- or voice- recognition
Detecting objects in images is an extremely important step in many image and video analysis applications. Object detection is considered as one of the main challenges in the field of computer vision, which focuses on identifying and locating objects of different classes in an image. The convolutional neural networks (CNNs) represent the heart of state-of-the-art object detection methods. They are used for extracting features. Several CNNs are available, for instance, AlexNet, VGGNet, and ResNet. These networks are mainly used for object classification task and have evaluated on some widely used benchmarks and datasets such as ImageNet
As the amount of video data generated tends to be pretty huge, with no way to handle and process all of it in a short span of time using manpower alone due to limitations in human capacity, video analytics is serving as a useful asset to make generated video data more valuable. Automated solutions, delivered by deep learning and artificial intelligence, can efficiently analyze the huge amount of data that videos generate, providing tremendously fast results.
Using AI in video analytics, a number of systems will be able to communicate with each helping in taking decisions and readily catching suspicious activities or predicting them before they can happen. There are some situations where a camera cannot take action due to some visual obstacle that is not included in camera tampering algorithms which means video analytics will not work. The situation can be beyond the line of sight.
Combining video analytics with other advanced technologies, including Real-time Location Systems (RTLS) or Radio-frequency identification Systems (RFID), can provide the exact data or location.
Facial expressions manifest not only emotions but also allied actions, behavioral patterns and give a lot of useful data when it comes to helping industries like Law enforcement, Forensics etc. Video analytics can be achieved based on data curation, sentiment analysis, and other advanced solutions. Expressions like “happy”, “sad”, “angry”, “scared”, “surprised” or “neutral” form the basis of video analytics.
An advanced video analytics solution may contain multiple functionalities and features including:
- People management: Crowd detection, queue management, people counting, people scattering, people tracking
- Vehicle management: Vehicle classification, traffic monitoring, license plate recognition, road data gathering
- Behavior monitoring: Motion detection, vandalism detection, face detection, privacy masking, suspicious activity detection
- Device protection: Protection against camera tampering, perimeter protection, intrusion detection, theft and threat detection
Artificial Intelligence(AI) for prevention of terrorist attacks
Identification of terrorist ideologies and prediction of future terrorist attacks have been proven to be of great importance and a time-consuming process. The planning of terrorism is very difficult to observe in terms of behavior. Therefore, a number of approximate signals are often used as a construct for the underlying variable planning of terrorism, such as possession of weapons, certain targeted keywords communication, watching certain YouTube videos, use of social networks or specific suspected phone numbers.
There are two practical analytics approaches to seeking to prevent terrorist attacks based on data. One is an inductive search for patterns in data, and the other is a deductive data search based on a model of criminal relations between suspicious persons, actions or things. Methodologically, this distinction is called pattern-based data mining versus subject-based data mining.
Only pattern-based data mining constitutes proper predictive analytics in which outliner segments are inductively identified in a multi-dimensional universe — segments identified by the algorithm as deviant in a number of central variables in which the deviation is then connected with terrorism. By contrast, subject-based data mining is merely traditional police investigation in a digitized form. No future predictions are involved, only the identification of a network of social relations. Searches in digitized data from larger and larger data volumes are performed more quickly based on the premise that crime is a social relation between people that leaves digital traces.
AI can be used to make predictions about terrorism based on communications metadata, financial transaction information, travel patterns and internet browsing activity, as well as publicly available information such as social media activity. The development and use of AI in the financial services sector has been spurred by mandatory reporting of suspicious activity in financial transactions.
In the case of counterterrorism, investigative approaches applied prior to an attack taking place are traditionally undertaken by the intelligence and security services. Broadly speaking, these approaches focus on working outwards from partially discovered plots or known suspects in order to find other involved parties or to identify links leading deeper into terrorist organizations.
Developments in AI have amplified the ability to conduct surveillance without being constrained by resources. Deep learning technologies also help analyze and process vast streams of footage. AI-driven text analysis could be used to ‘understand’ the content of private messages without requiring the attention of a human analyst.
Facial recognition technology, for instance, may enable the complete automation of surveillance using CCTV in public places in the near future. A well-trained deep learning solution allow Intelligent video analytics to analyze facial data more quickly by providing more accurate face detection with faster response time, thus creating a powerful method for facial recognition.
The promise of more sophisticated AI is that it will be able to make better predictions based on behaviours in a range of different areas, rather than by crude profiling. As terrorist organizations conduct more activity online, research has shown that it might be possible to use AI to analyse communications and discern characteristics such as degree of radicalization, aggressive intent, or the genesis of terrorist movements, and even to predict the incidence of violent attacks. Accuracy of predictive models based on one source of data can be progressively improved by integrating the results from other feeds.
Machine learning for Predicting Terror Activity Before It Happen
The patterns of attacks planned and carried out by terrorists may seem random on the surface, but in fact, they are typically organized and premeditated actions chosen carefully and deliberately. Moreover, attacks by the same organizations and individuals tend to be substantially related in terms of certain distinguishable characteristics. Therefore, there must be some patterns or informal rules guiding the activities of terrorist organizations. After analyzing these characteristic patterns of activity by terrorist organizations, authorities can make more detailed predictions and analyses of terrorist organizations to attack them more accurately and increase the time available for the prevention and prediction of terrorist attacks.
In Oct 2019, Data scientists have reported to have developed an early-warning model that can successfully predict how lethal a terror organization will become in the future based on only its first 10 attacks. The predictive model, developed by researchers in Northwestern University’s Kellogg School of Management, will allow security forces to better identify and target the most destructive groups to potentially stop them before they grow too powerful.
“This early warning is huge because not only can it help the government target and neutralize the groups with the most potential for destruction, it also can help the government strategically deploy resources and avoid spending billions of dollars fighting a group that is likely to burn out on its own anyway,” said Brian Uzzi, corresponding author on the study and the Richard L. Thomas Professor of Leadership and Organizational Change at Kellogg.
The model, which utilizes data publicly available through the Global Terror Database (GTD) and the RAND Database of Worldwide Terrorism Incidents (RDWTI), has considerable power in predicting a terror organization’s lifetime violence after just 10 terror attacks. The National Consortium for the Study of Terrorism and Responses to Terrorism (START) has prepared a dataset known as Global Terrorism Database (GTD) (https://www.start.umd.edu/gtd). GTD contains information about terrorist activities from 1970 until 2018, including more than 181,000 different instances of terrorism.
The GTD provides researchers with comprehensive, reliable, and open-source data, in which there are many potential correlations and patterns to be found. Mining and identifying these patterns using digitally driven methods has become a research topic in the field of informatics.
Researchers calibrated the model using data from terror groups operating between the years of 1970 and 2014. They noted that some of the most interesting model predictions were for groups that operated with very few attacks in the beginning, only becoming lethal much later. Among these groups were United Liberation Front of Assam, Al-Shabaab and Moro Islamic Liberation Front. “The model can predict the future impact of some of these sleeper groups even while they are still operating in an under-the-radar way,” said study co-lead author Yang Yang, a postdoctoral fellow at Northwestern.
The AI methods use machine learning to build models based on data, and then make inferences from those models. Large amounts of effort, particularly from the academic community, have been devoted to developing models that predict the location and timing of terrorist attacks. Basic approaches have incorporated the ‘aftershock effect’, whereby the chance of another event is increased in the wake of an attack (a phenomenon also observed with crimes such as burglary) to make surprisingly accurate predictions about terrorist attacks. Other approaches have predicted the impact of external factors – such as political conditions – on the incidence of attacks.
“Previous models, for the most part, are useful for understanding the context in which terror activity is likely to take place, but they are too confined to the locale and not useful in predicting individual organizations’ behavior,” said Adam Pah, co-lead author of the study and clinical assistant professor of management and organizations at Kellogg.
“Essentially we said, ‘What if we think of terror organizations like a business whose product is lethality? How do we predict their success in producing that product?’” Uzzi said. 61 terror groupsemerged each year from 2000 to 2015, leading to an 800% increase in global terror attacks
Venture capitalists and business investors routinely use publicly reported information such as cash flows and technological skills to predict company behavior and success. Such information is not available for secretive terror organizations, so the researchers worked to develop proxies based on observable behavior.
For example, business investors often view the timing of a company’s product releases as a proxy for resources. They assume that a company that regularly launches new products likely has more resources than a company that delivers new products at random. Similarly, the researchers’ model uses the timing of attacks as a proxy for a terror group’s resources and organizational strength. The researchers were able to confirm these notions with factors like the diversity of weapons used, the sophistication of those weapons, and their attack capabilities, defined as the extent to which the group succeeded in carrying out the mission of the attack.
Predictive model draws on business forecasting methods to predict the lifetime lethality of terror organizations.Researchers noted that Islamic State had extraordinarily strong attack capabilities near the 90th percentile of all terror groups in similar ages, even though the group demonstrated an irregular attack cadence that was indicative of unstable resources. After just 10 attacks, the model placed ISIS among the terror groups with the most potential for committing exceptionally deadly attacks.
Uzzi, Yang and Pah all are members of the University’s Northwestern Institute on Complex Systems (NICO) and are faculty at the Kellogg School of Management and the McCormick School of Engineering.This work was supported by the U.S. Army Research Laboratory and U.S. Army Research Office (grant W911NF-15-1-0577) as well as Army Research Laboratory Network Science CTA (under Cooperative Agreement W911NF-09-2-0053)
Machine learning methods have been recently explored to develop techniques for counterterrorism based on artificial intelligence (AI). The advancements in computer technologies have been able to create much powerful computer systems to perform the required computation in DNN.
In the 2020 paper by M. Irfan Uddin et al., NN and DNN models are used to make predictions of different factors that lead to terrorist activities. The model is helpful for law enforcement agencies to make predictions before an incident actually happens and potentially causes the loss of precious lives. The predicted factors are explained below.
(i) Suicide: to predict whether a terrorist activity is going to be suicide or not.
(ii) Success: to predict whether a terrorist activity will succeed or not.
(iii) Weapon type: to make a classification of the general type of weapons used in terrorist activity.
(iv) Region: to classify the region that will be targeted by the terrorist activity.
(v) Attack type: to classify the type of attack carried out as a terrorist activity.
These predictions are important to understand in order to perform counterterrorism. Deep learning can make these predictions efficiently and can help law enforcement agencies to devise mechanisms to deal with terrorists and protect the lives of individuals. With the help of these tools, a terrorist activity can be stopped before it can actually happen and make destructions in terms of lives, infrastructure, or law.
In 2021, Gian Maria Campedelli study aimed at bridging artifcial intelligence and terrorism research by proposing a new computational framework based on meta-graphs, time-series, and forecasting algorithms commonly used in the felds of the machine and deep learning. They explain, “Retrieving event data from the Global Terrorism Database, we focus on all the attacks that occurred in Afghanistan and Iraq from 2001 to 2018 and construct 2-day-based meta-graphs representing the operational connections emerging from three event dimensions: utilized weapons, deployed tactics, and chosen targets. Once meta-graphs are created, we derive time-series mapping the centrality of each
feature in each dimension. Te generated time series are then utilized to learn the existing recurring patterns between operational features to forecast the next most likely central – and therefore popular – targets.”
The problems include the opportunity cost of false positives/false negatives, the statistical quality of the prediction and the self-reinforcing, corrupting recursive effects of predictive analytics, since the method lacks an inner meta-model for its own learning- and pattern-dependent adaptation. The conclusion is algorithms don’t work for detecting terrorism and is ineffective, risky and inappropriate, with potentially 100,000 false positives for every real terrorist that the algorithm finds.
Social media are interactive computer-mediated technology that facilitates the sharing of information via virtual communities and networks. And Twitter is one of the most popular social media for social interaction and microblogging. An approach was proposed to build a dictionary using tweets containing hashtags like Al-Qaeda, Jihad, Terrorism, and Extremism and by collecting the relevant words. Another approach using ISIS related tweets to predict the future support was developed, where twitter data is used to study the antecedents of ISIS support of users. Go et al. have done another study in tweets sentiment classification.
In 2020, Researchers from Bangladesh led by Aditi Sarker developed improved system model to analyze twitter data and detect terrorist attack event. The main objectives of this paper are to filter of extracting for finding the words and geo-location, to introduce Pattern Matching and Weight Assigning (PMWA) machine for finding the weights of filtered words and to calculate the weighted sum of tweeted words in linear time complexity by applying Aho-Corasick automata. “Based on paper which was mainly developed to get detecting phase during natural disaster, in this paper we try to develop an improved technique for detection of terrorist attack event and we applied Aho-Corasick algorithm to perform pattern matching so that we can assign weights to extracted tweeted words, which was not applied previously.”
US law enforcement
Earlier, it came to light that law enforcement agencies had begun testing pilots with Amazon Rekognition, the company’s cloud-based facial recognition technology. According to confidential documents obtained by the Intercept, IBM has developed object recognition technology capable of identifying people by physical characteristics like skin color with help from the New York Police Department.
With secret access to NYPD-captured videos, IBM developed this technology to allow police to search camera footage for persons with a specific hair color, facial hair type, or skin color, the Intercept reports. The development of this technology dates back to 2012 when counterterrorism officials gained access to skin color-searching capabilities. The technology originally focused on object recognition specifically but was later honed to identify persons by age, “head color,” gender, and skin color. Some of the IBM researchers had qualms about the technology if its accuracy improved in the future. NYPD told the Intercept that its collaboration with IBM was about “finding a way to shorten the time to catch the bad people” after a crime.
Speaking at AWS re:Invent 2018 in Las Vegas this week, Deputy Assistant Director Christine Halvorsen explained that the FBI moved its counterterrorism data to the cloud on AWS after the deadly shooting, which “resulted in a 98 percent reduction in manual work for analysts and 70 percent cost reductions,” fedscoop reported. “We had agents and analysts, eight per shift, working 24/7 for three weeks going through the video footage of everywhere Stephen Paddock was the month leading up to him coming and doing the shooting,” said Halvorsen, who today was named “Person of the Year” by Homeland Security Today. “If we had loaded that up into the cloud, the estimate is it would’ve taken us a day using Amazon Rekognition to recognize where he was in the videos. That’s all we were trying to do: narrow down where in the videos he was and who he was meeting with to make sure there wasn’t anybody else part of the conspiracy,” she added.
However, a study by the ACLU found performance flaws in Rekognition leading to its incorrectly matching 28 members of Congress, identifying them as other people who have been arrested for a crime. The members of Congress who were falsely matched with the mugshot database used in the test included Republicans and Democrats, men and women, and legislators of all ages, from all across the country. Nearly 40% of Rekognition’s false matches in the test were of people of color, even though they made up only 20% of Congress.
China’s city surveillance programme
China’s city surveillance programme has been driven by government policies and initiatives, including the 2005 Skynet Program to strengthen public security by installing cameras in key public areas, the completion of the installation of cameras in all key public places by 2020, upgrading existing cameras to HD resolution and ensuring all video footage from these areas is accessible to the authorities. The Xue Liang program, launched in 2016, aims to connect all cameras installed in villages, towns and districts to a central surveillance platform from county to national level, and to share video across police forces, emergency services and other government agencies. The scale of these surveillance operations means that users need to find ways of interpreting and processing the vast amounts of data produced – hence the drive for deep learning technologies on the part of Chinese manufacturers. Video analytics based on deep learning use a set of algorithms to enable systems to ‘learn’ from examples unsupervised or semi-supervised, and then apply that learning to future scenarios.
Chinese startup SenseTime, which makes AI-powered surveillance software for the country’s police, and which received in April 2018 a new round of funding worth $600 million. This funding, led by retailing giant Alibaba, reportedly gives SenseTime a total valuation of more than $4.5 billion, making it the most valuable AI startup in the world, according to analyst firm CB Insights. Most notably, SenseTime also outfits Chinese law enforcement with facial recognition and tracking services. For example, the company says that software it provides for the security bureau of Guangzhou (one of China’s three biggest cities with a metropolitan population of around 25 million) is used to match surveillance footage from crime scenes to photos from a criminal database, and has identified more than 2,000 suspects and solved “nearly 100 cases.”
Global Intelligent Video Analytics Market growth
Current law enforcement systems are increasingly unable to cope with the sheer volume of surveillance material captured and stored every day. This is only set to rise, with the population of video cameras increasing by at least 12% per year. These video streams will only ever be useful if processes to search and analyze the mountain of data keep pace. As it stands today vital information is missed because the vast majority of the video is simply never viewed.
MarketsandMarkets estimates the global video analytics market is expected to grow from USD 4.9 billion in 2020 to USD 11.7 billion by 2025, at a Compound Annual Growth Rate (CAGR) of 19.0% during the forecast period.
Video Analytics is the technology, software and process of using computers to automatically analyze video footage from a security camera. Using video analytics makes your surveillance system more efficient, reduces the workload on security and management staff, and helps you capture the full value of security video by making your IP camera system more intelligent in its work. Benefits of video analytics includes – high accuracy, license plate recognition, strange object recognition, auto-tracking, ease of implementation, reduced labor costs, improves security, and real-time surveillance.
The researchers have defined AI Video Analytics as a solution that is running deep learning algorithms on a platform that is most likely to be built on a GPU chip architecture. These solutions are very much in their embryonic stage and the researchers best estimate is that global sales of AI Video Analytic solutions in 2017 was only around $115 million, much of this being installed in China on Safe City projects.
China is the largest and fastest growing market for video surveillance – the domestic market is forecast to be worth up to $20bn in 2018. The adoption of video surveillance systems in China is booming, driven by the country’s safe cities initiative (mostly by the public sector) and a burgeoning private sector. There are huge surveillance installations in towns and cities – some complete with face recognition, behaviour analysis and ANPR. In the southwestern city of Guiyang, for example, images of all its 3.5 million residents are held by the authorities, and can be captured and detected on cameras equipped with face recognition. Cameras can also be used to estimate the age, gender and ethnicity of subjects.
The scale of these surveillance operations means that users need to find ways of interpreting and processing the vast amounts of data produced – hence the drive for deep learning technologies on the part of Chinese manufacturers. Video analytics based on deep learning use a set of algorithms to enable systems to ‘learn’ from examples unsupervised or semi-supervised, and then apply that learning to future scenarios. The first installations in China of video surveillance equipment based on deep learning took place in 2016.
This growth is due in large part to major advances in semiconductor architecture, which is enabling much faster processing; Empowering deep learning and machine learning algorithms to analyze data many times faster than was previously possible. Venture capitalists are now pouring billions of dollars into financing Artificial Intelligence (AI) chip and analytic software companies. Indeed while researching this report, the researchers identified 128 companies across the world that are now in some way helping (hardware & software) to deliver AI video analytic solutions.
Nvidia has emerged as the early leader in AI chips and is particularly strong in video analytics. Nvidia’s edge is that its PC gaming processors (GPUs) can be scaled up to handle AI software, thanks to their parallel processing circuitry which can handle complex multiple tasks.
The major players in the market are IBM Corporation, Intellivision Technologies Private Ltd., Honeywell International Incorporation, , Puretech Systems Inc., Axis Communications, I2V Systems Private Ltd., Qognify, Intuvision, Inc., Genetec Inc., Aventura Technologies Inc., Allgovision Technologies Pvt. Ltd. and Avigilon Corporation. Other players are Viseum International (Potters Bar, England), AllGoVision (Karnataka, India), Huawei Technologies Co., Ltd. (Shenzhen, China), Agent VI (New York, United States), Gorilla Technology Group (Taipei, Taiwan), Cisco Systems, Inc. (California, United States), IBM Corporation (New York, United States), Kiwisecurity (Vienna, Austria), Axis Communications AB (Lund, Sweden), Robert Bosch GmbH (Stuttgart, Germany), Honeywell International Inc. (Charlotte, United States), Motorola Solutions, Inc. (Avigilon Corporation) (Illinois, United States), ULTINOUS Zrt. (Budapest, Hungary).
There is still much to be done in perfecting the technology and getting it to market, but these new tools’ have opened up the opportunity to bring AI products to the video analytics market potentially revolutionizing its performance and capability. And if it can deliver, it will further drive demand for intelligent video surveillance, not just for new projects but open up a vast latent potential for retrofitting millions of existing camera installations.