Social Perspective of Suspicious Activity Detection in Facial Analysis: An ML-Based Approach for the Indian Perspective

Suspicious Activity

The chapter describes initially the basics of initial crime status at world, face recognition and its tells and techniques, machine learning and data analytics, and its tool and techniques. For ease of audiences, it also gives a brief review of Data mining, Regression, AI, and CPS. In the next section of the literature review, the famous authors and the gist of their work on the content have been enlisted. Then in the application section, the latest classification techniques of support SVM, Dlib, CNN, and RNN has been introduced and their application on the data set is reflected. In the result section, the different emotions and objects have been recognized and their accuracy has been discussed. Then in the later part the recommendations, novelty, application, limitations have of the research work are explained followed by concluding remarks.

Face Recognition

The world is going through an extraordinary heightening of the CPS also known as cyber-physical systems, which are envisaged to transfigure our world by creating new software applications and services in a large variety of domains such as intelligent transportation systems, systems for monitoring the environment, and intelligent transportation systems and so on. The ICT sector also known as the information and communication technology sector is also going through an impact making growth in data traffic, which is driven by the high usage of tablets, smart cellphones, and video streaming along with the notable growth of deployment of sensors that are forecasted soon [Toygar et al., 2003; Viola et al., 2004].

Machine Learning

If the performance at tasks as measured by the performance measure (known as P) improves, then the agent is said to have to gain knowledge or learning from experience denoted by E for some set of tasks denoted by T. For example, Mailing System [Mitchell, 1997], Playing Checkers game.

There are different categories of machine learning:

• 1.

  Supervised machine learning teaches us about learning an input and output map (classification: categorical output, regression: continuous output)

• 2.

  Unsupervised machine learning teaches us about discovering patterns in the input dataset (clustering: cohesive grouping, --association: frequent co-occurrence)

• 3.

  Reinforcement machine learning – learning control

Data Analysis

This is the technique used for extracting useful, relevant, and meaningful information from the huge amount of data in a structured manner for fulfilling the purpose of Parameter estimation also known as inferring the unknowns, identification and classification of features, testing of Hypothesis, and detection of fault and Development and Prediction of Model also known as Forecasting.

Tools of Data Analysis: Weka, R, Python

Python is a high-level language, which is also object-oriented and is used highly in the problems of semantic dynamic and also used for general-purpose programming. It is interpreted programming language. It is used for: web development (server-side), software development. The way to run a python file is like this on the command line:

• helloworld.py

• print (“Hello, World!”)

Weka and R

It is a freely available s/w package containing a collection of machine learning algorithms under the GNU (General Public License). It is an open-source. The algorithms present in Weka are all coded in java and they can be used by calling them from their java pod. However with also provides a graphical user interface from which the algorithms can directly be applied to data sets.

R software: R is the programming language. It is freely available s/w. It is used for, statistical and analysis, data manipulation, graphic display. Effective data handling and storage of o/p is possible.

• > 2 + 2=4 [Tripathi et al., 2014].