Intelligent Criminology Lab

Principal Investigator:

Dr. Muhammad Usman Ghani Khan
Al-Khawarizmi Institute of Computer Science
University of Engineering and Technology,
G. T. Road, Lahore

Phone: +92-42-99250229 Email: usman.ghani@kics.edu.pk

 

CO-Principal Investigators:

Contact Information Co-PI(1)

Name: Prof. Dr. Waqar Mahmood

Al-Khawarizmi Institute of Computer Science
University of Engineering and Technology,
G. T. Road, Lahore

Email: director@kics.edu.pk

Mobile: 0321-4755919

Contact Information Co-PI(2)

Name: Dr. Sheikh Faisal Rasheed

Al-Khawarizmi Institute of Computer Science
University of Engineering and Technology,
G. T. Road, Lahore

Email: faisal.rashid@kics.edu.pk

Mobile: 03062129892

Contact Information Co-PI (3)

Name: Dr. Muhammad Aslam

Department of Computer Science & Engineering,
University of Engineering and Technology,
G. T. Road, Lahore

Email: maslam@uet.edu.pk

Mobile: 0092 – 42 111-128-128 Ext: 237 

Contact Information Co-PI (4)

Name: Prof. Dr. Muhammad Tahir

Al-Khawarizmi Institute of Computer Science
University of Engineering and Technology,
G. T. Road, Lahore

Email: haroon.mahmood@nu.edu.pk

Mobile: 0092 – 42 111-128-128 Ext: 237

 

Goal: Application of AI to prevent and mitigate crimes in Pakistan

Scientific Objectives:

  • To develop an efficient crime and security monitoring system tailored for Pakistani community
  • To provide solution to equip government agencies to deal with criminal activities.
  • To identify potential culprits and dissuade criminals from committing crimes using in house image processing, computer graphics and text analytics capabilities.
  • To produce compact and comprehensive textual repository of video sequences for storage, indexing and retrieval tasks
  • To extract crime patterns and inform law enforcement personnel by analyzing the audio, visual and textual data
  • To predict human behavior, detect criminal activity and generate alarms for the deterrence of crime
  • To preempt security agencies and police department to monitor specific area for suspicious activity or for suspects.
  • To promptly respond to an emergency situation for better handling of security threats
  • To analyze crowd for making predictions for potential behaviors like sudden dispersion, growth, restlessness etc.
  • To learn and extract behavioral patterns for each considered crowd types to tackle with upcoming unrest situations.
  • To develop abnormal situation handling simulations on the basis of generated analysis of audio, visual and textual streams.
  • To create complete application for behavioral mining of crowd with respect to all aspects mentioned above.

Description of Lab:

Terrorism/Criminal intent is one of the alarming threats that have shaken nations and challenged their existence. Pakistan is among those countries who have suffered the most. In spite of its best efforts Pakistan is unable to stop this wave of terrorism. Criminology is related to study of crimes, their causes and expansions, monitoring and understanding behavior of criminals, generating models of criminal intents and then predicting future crimes. This Laboratory will use multimodal criminal data i.e., textual, audio and video, to develop novel methods to incorporate intelligence in criminology concepts and generate solutions to related indigenous problems. We use our technical developments in a variety of real-world applications, including:

  • Intelligent analysis of textual information from different sources for the purpose of identifying crime patterns, series and trends.
  • Design and maintain criminal intelligence databases for national security and law enforcement action.
  • Video Surveillance to map crime locations and analyze crime hotspots.
  • Develop intelligent systems for computational forensics to examine video, audio, textual or any other type of evidence.
  • Intelligent automated call-analysis agents to analyze phone call records for certain phrases.
  • Intelligent systems to monitor Internet traffic.
  • Intelligent monitoring and surveillance of road traffic.
  • Intelligent Surveillance agents to analyze social networks based data over Internet to identify the hotspot of future disorder or crime.

Through research and development in this laboratory the aim is to play active role in solving problems at local and national level in areas of law enforcement, public/national safety and crime control

 Applications Domains:

Our broad scientific area is Intelligent criminology where we will be using algorithms from Artificial Intelligence for tackling the versatile problems and issues in the field of criminology. Some of the examples of this lab’s work relate to the use of multi-modal data (audio, video and

textual) to identify criminals, suspects. crime scene identification, criminal behavior modeling and predictions. Overall, we are focusing on four application domains.

  1. Information Retrieval (IR) framework for retrieval of criminology related visual data
  2. Criminal behavior modeling and prediction from social media contents and
  3. Forensic Animation: A 3D Visualization of Crime Scene simulation, monitoring and training of law enforcers.
  4. Crowd Analytics: Behavioral Mining for Visual Crowds and Individuals Tracking