New and Improved Deep Learning-Supported Behavioural Analytics – Wavesys Global’s Behavior Analysis Systems for Law Enforcement

New Delhi [India], July 13: Wavesys Global recently released the latest version of Behavioral Analytics (v2.1.0), which introduces a new WVCAbehaviour license with several new algorithms. These algorithms offer alerts on behavior/ motion such as fighting, aggressive behavior, slip and fall, trip and fall, possession of hand-held objects, etc., using Deep Learning modules.
Talking about this new version, Senior Regional Director of Wavesys Global, Abhishek Kumar, said, “Experts in human psychology believe that most humans behave differently when no one’s looking; habits and activities that seem legally, morally, ethically, or behaviorally  ‘bad’ or ‘unacceptable’ crawl out when nobody watches. And most of the time, nobody is watching, even when there is a CCTV camera. Behavior Analytics is just the solution to this ‘unattended footage’ situation. Our new WVCA behavior license can bring breakthrough resolutions in a lot of criminal cases and can even help in avoiding some crimes altogether.”

Explaining the latest development of this new offering, he further added, “Can you imagine the sort of advantage the Law Enforcement departments can have if they have access to a system that has Deep Learning-supported trackers—the skeleton tracker providing 18 body points of interest, the fall tracker, the hand-held object tracker, and the ability to detect aggression and fighting? It is, indeed, a game changer.

The WVCA Behaviour license goes beyond the standard use cases covered under the previous WVCAproAI license. It consists of a wide variety of behavior detection algorithms, such as the Hand Object Interaction Tracker and Aggressive Behaviour Rule, and includes the reworked Deep Learning Skeleton Tracker (DLST) and Fall Detection Rule.

The redeveloped and updated Deep Learning People Tracker (DLPT) improves detection range and performance in clustered and busy scenarios by utilising head and upper body detection. It offers highly accurate detection, tracking, and counting, even in complex or crowded environments.

The Aggressive Behaviour Detection feature detects and analyses the field of view over a set number of frames and creates alerts wherever it notices that a fight or any kind of aggressive behavior is likely to occur.

The new Hand-Object Interaction (HOI) Tracker algorithm detects hands and objects held in them. It enables the use of all WVCA-tracked objects, zones, and rules to create advanced behavior detection rules, which can be beneficial in several different circumstances, including cases like retail shelf protection and the prevention of shrinkage at self-checkout terminals.

Additionally, there are also numerous quality-of-life improvements that the new version offers. The new version is capable of generating events on its own if the input rule is triggered a definite number of times over a set period; it can even counter-reset to zero. The UI support for SSL and Certificates enables a self-managed end-to-end encrypted connection between the browser and the back-end services. And ONVIF integration allows external applications to utilize the events triggered by the set rules.

A MEDLEY OF OBJECT RECOGNITION AND BEHAVIOR ANALYSIS

Behavioral Analytics can be of prime importance in any and every industry due to its vast and versatile application possibilities. However, the latest version of WVCA Behavioral Analytics has been created to bring great benefit to Law Enforcement and Security in particular.

The greatest challenge faced by law enforcement departments is keeping a watch on everyone, everywhere. WVCA Behavioral Analytics tools can be of immense value in accomplishing this objective to a large extent.

The Deep Learning People Tracker (DLPT) and Deep Learning Skeleton Tracker (DLST) work towards calculating body count and body point movements, thus tracking and creating alerts in case of possible vandalism, fights, suspicious package dropping, crowd control, and so on.

On the other hand, the Hand-Object Interaction (HOI) Tracker helps in the identification of weapons or objects of suspicion or interest.

Aggressive Behavior Detection aligned with DLST provides alerts when signs of fights, vandalism, or other violent behaviour are detected.

Wavesys Global specialises in Video Content Analytics, Video Management Software, IP Cameras, Facial recognition systems, Forensic search, IP Access Control solutions, and IT Hardware- Industrial Grade Switches. Waveys Global offers a complete analytics suite with features like speed violation, counting, people tracker, zone and line classification, object classification, pattern recognition, post estimation, and many more.