Machine Learning Integration into Police Surveillance Footage for Crime Detection

Author #1

Keywords:

  1. Artificial Intelligence (AI): the capability of a machine to imitate intelligent human behavior. (Merriam Webster, 2019)

  2. Crime-Index: the eight primary crimes that the FBI collects which are Murder, Forcible-Rape, Robbery, Aggravated-Assault, Arson, Burglary, Larceny-Theft, and Motor-Vehicle-Theft. (Wallace, 2018)

  3. Facial Recognition: an advanced technology that helps in discerning and identifying human faces from an image or video. (Kanchwala, 2019)

  4. Machine-Learning (ML): the process by which a computer can improve its performance (as in analyzing image files) by continuously incorporating new data into an existing statistical model. (Merriam Webster, 2019)

  5. Predictive Analytics: a category of data analytics aimed at making predictions about future outcomes based on historical data and analytics techniques such as statistical modeling and machine learning. (Edwards, 2019)

References:

  1. Machine Learning. (n.d.). Retrieved from https://www.merriam-webster.com/dictionary/machine learning

  2. Artificial Intelligence. (n.d.). Retrieved from https://www.merriam-webster.com/dictionary/artificial intelligence

  3. Kanchwala, H. (2019, April 29). Facial Recognition: Definition, History, Working, and Applications. Retrieved from https://www.scienceabc.com/innovation/facial-recognition-works.html

  4. Edwards, J. (2019, August 16). Predictive analytics: Transforming data into future insights. Retrieved from https://www.cio.com/article/3273114/what-is-predictive-analytics-transforming-data-into-future-insights.html

  5. Vincent, J. (2018, January 23). Artificial intelligence is going to supercharge surveillance. Retrieved from https://www.theverge.com/2018/1/23/16907238/artificial-intelligence-surveillance-cameras-security

  6. Aravindan, A. (2018, April 13). Singapore to test facial recognition on lampposts, stoking privacy... Retrieved from https://www.reuters.com/article/us-singapore-surveillance/singapore-to-test-facial-recognition-on-lampposts-stoking-privacy-fears-idUSKBN1HK0RV

  7. Doffman, Z. (2019, August 27). Hong Kong Exposes Both Sides Of China's Relentless Facial Recognition Machine. Retrieved from https://www.forbes.com/sites/zakdoffman/2019/08/26/hong-kong-exposes-both-sides-of-chinas-relentless-facial-recognition-machine/#78de41942b74

  8. Shouk, A. A. (2018, November 7). Dubai CCTV cameras to use AI, face recognition. Retrieved from https://gulfnews.com/uae/government/dubai-cctv-cameras-to-use-ai-face-recognition-1.2163726

  9. Wallace, L. (2018, June 25). What is the Crime Index? Retrieved from https://www.legalmatch.com/law-library/article/what-is-the-crime-index.html

  10. Chicago, Illinois. (n.d.). Retrieved from http://www.city-data.com/city/Chicago-Illinois.html

  11. Crime rate in New York, New York (NY): murders, rapes, robberies, assaults, burglaries, thefts, auto thefts, arson, law enforcement employees, police officers, crime map. (n.d.). Retrieved from http://www.city-data.com/crime/crime-New-York-New-York.html

  12. Los Angeles, California. (n.d.). Retrieved from http://www.city-data.com/city/Los-Angeles-California.html

  13. Crime rate in San Francisco, California (CA): murders, rapes, robberies, assaults, burglaries, thefts, auto thefts, arson, law enforcement employees, police officers, crime map. (n.d.). Retrieved from http://www.city-data.com/crime/crime-San-Francisco-California.html

  14. Atlanta, Georgia. (n.d.). Retrieved from http://www.city-data.com/city/Atlanta-Georgia.html

  15. Raj, B. (2018, August 10). How to Automate Surveillance Easily with Deep Learning. Retrieved from https://medium.com/nanonets/how-to-automate-surveillance-easily-with-deep-learning-4eb4fa0cd68d

Abstract

Machine Learning (ML) is now starting to be used in surveillance cameras as the technology to aid crime-detection in major cities. For example, this technology has been proven useful in major cities in the eastern hemisphere such as Tokyo (Vincient, 2018), Singapore (Aravindan, 2018), Hong-Kong (Doffman, 2019), and Dubai (Shouk, 2018). For instance in Singapore (Aravindan, 2018), Hong-Kong (Doffman, 2019), and Dubai (Shouk, 2018) use facial-recognition, while Tokyo (Vincient, 2018) uses predictive-analytics. The purpose of this research is to explore the possible effects of crime-detection of this technology in major cities in the United States using the city of Atlanta as a case study. Atlanta can be considered as a representative city due to the amount and variety of crimes, which are similar to other major cities in America. For instance, the average 2017 crime-index for major cities, Chicago (City-Data, 2017), New-York-City (City-Data, 2017), Los-Angeles (City-Data, 2017), and San-Francisco (City-Data, 2017), in the US is around 500 out of 100,000 and Atlanta’s 2017 crime-index is 520.8 (City-Data, 2017) which proves the fact that Atlanta is a great case study. We used footage from the Atlanta-Police-Department (ATL-PD) and Google-Images which are tagged to a specific type of crime. We adapted a machine learning model from TensorFlow-Object-Detection-API (Raj, 2018). Then we trained the model with the police footage to differentiate what type of crime has occurred if one has occurred. Then we used part of the police footage, separated for testing, to test the accuracy of the model. Our results show that the model is performing as accurate and fast, if not better, as the current methods used by the ATL-PD. Once integrated into the ATL-PD surveillance-system, this method will aid crime-detection, eventually progressing to crime-prevention.

 
Nov 2nd, 8:00 AM Nov 2nd, 9:00 AM

Machine Learning Integration into Police Surveillance Footage for Crime Detection

Nesbitt 2201

Machine Learning (ML) is now starting to be used in surveillance cameras as the technology to aid crime-detection in major cities. For example, this technology has been proven useful in major cities in the eastern hemisphere such as Tokyo (Vincient, 2018), Singapore (Aravindan, 2018), Hong-Kong (Doffman, 2019), and Dubai (Shouk, 2018). For instance in Singapore (Aravindan, 2018), Hong-Kong (Doffman, 2019), and Dubai (Shouk, 2018) use facial-recognition, while Tokyo (Vincient, 2018) uses predictive-analytics. The purpose of this research is to explore the possible effects of crime-detection of this technology in major cities in the United States using the city of Atlanta as a case study. Atlanta can be considered as a representative city due to the amount and variety of crimes, which are similar to other major cities in America. For instance, the average 2017 crime-index for major cities, Chicago (City-Data, 2017), New-York-City (City-Data, 2017), Los-Angeles (City-Data, 2017), and San-Francisco (City-Data, 2017), in the US is around 500 out of 100,000 and Atlanta’s 2017 crime-index is 520.8 (City-Data, 2017) which proves the fact that Atlanta is a great case study. We used footage from the Atlanta-Police-Department (ATL-PD) and Google-Images which are tagged to a specific type of crime. We adapted a machine learning model from TensorFlow-Object-Detection-API (Raj, 2018). Then we trained the model with the police footage to differentiate what type of crime has occurred if one has occurred. Then we used part of the police footage, separated for testing, to test the accuracy of the model. Our results show that the model is performing as accurate and fast, if not better, as the current methods used by the ATL-PD. Once integrated into the ATL-PD surveillance-system, this method will aid crime-detection, eventually progressing to crime-prevention.