Working Women Alert By Using Predictive Analysis

The main focus of our project is women safety to safe the women from the unpredictable behavior of unknown person. Basically, it can work through image processing. The image processing can detect the user expression. By using predictive analysis technique and trained CNN model we predict the behavio

Project Title

Working Women Alert By Using Predictive Analysis

Project Area of Specialization

Artificial Intelligence

Project Summary

The main focus of our project is women safety to safe the women from the unpredictable behavior of unknown person. Basically, it can work through image processing. The image processing can detect the user expression. By using predictive analysis technique and trained CNN model we predict the behavior of user and send alert message. If user can be harassed by the unknown person and user expression should be change then alert message/ alarm should be beep in security office. This way we save the working women for unpredictable behavior for unknown person.

Project Objectives

• Conduct research on different approaches for emotion recognition in the wild.
• To ensure that women are safe in working place.
• Design and implement one approach and perform experiments on various aspects of the chosen approach to improve it further.
• Carry out performance analysis of the different methods and models.
• Develop a web app and an accompanying Android app, allowing users to upload an image for emotion recognition from their PC or taken immediately from the camera of their smartphone/tablet. Also, ensure that the user interface for both the desktop and mobile applications is easy to use.

Project Implementation Method

The module consists of the following four steps –

1. Face detection

2. Feature pre-processing

3. Individual facial emotion CNN forward pass

4. Emotion Recognition

The first step of the bottom-up module is face detection which was achieved using a combination of HoG face detection algorithm and Multi-task Cascaded Convolutional Networks (MTCNN) algorithm. This step returns a list of coordinates of all the faces detected in the image. In the second step, the features obtained from the first step are cropped, scaled, normalized and aligned. In the third step, the preprocessed features are fed into a pre-trained CNN model (trained on isolated facial images and an array containing the individual emotion predictions for each face is obtained. Finally, an average of all the array elements is computed in order to get the overall emotion prediction.

Benefits of the Project

Detecting emotions with technology is quite a challenging task, yet one where machine learning algorithms have shown great promise. By using Facial Emotion Recognition, businesses can process images, and videos in real-time for monitoring video feeds or automating video analytics, thus saving costs and making life better for their users. And this way we secure the working place for women’s.

Technical Details of Final Deliverable

The project hardware deliverables can be used in technical deliverables is:

• Camera
• Workstation (Hard Disk min 1TB, RAM min 8GB etc.)
• Alarm

Machine configuration –

• 12 GB RAM
• Nvidia Tesla K80 GPU

 The project software deliverables can be used in technical deliverables is:

• Python
• CNN
• Asp.net
• Android Studio
• My SQL

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Security

Other Industries

Core Technology

Artificial Intelligence(AI)

Other Technologies

Sustainable Development Goals

Gender Equality

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com