Virtual gym instructor

A mirror is use to see reflection whereas a smart mirror to see reflection as well as calendar, remainder, news, weather forecasting and traffic flow. The Internet of Things refers to the billions of physical devices around the world that are now connected to the internet, all collecting and sharing

Project Title

Virtual gym instructor

Project Area of Specialization

Electrical/Electronic Engineering

Project Summary

A mirror is use to see reflection whereas a smart mirror to see reflection as well as calendar, remainder, news, weather forecasting and traffic flow. The Internet of Things refers to the billions of physical devices around the world that are now connected to the internet, all collecting and sharing data. It is possible because of advancement in technology, cheap hardware (i.e. ESP8266, Ri3) and their connectivity with each other. We will make a smart mirror, which will have all these qualities as well as it will also train, the members of gym. In our proposed project, we are providing an efficient solution to a very common problem which is discussed in problem statement; we just require internet access, microcontroller and display to make smart gym. First when member will arrive, there will be a face detection for login and the mirror will automatically display its schedule of workout. Virtual trainer will guide the member with routine tasks like which exercise you have to perform today, the correct way to perform them by showing a demo video and it will keep on monitoring if you are not doing it correctly or not.

Project Objectives

The main objectives of our project are:

1)Detect the user with face recognition

2)Show daily routine of gym

3)Display a demo of exercise

4)Monitor the workout and guide if member is not performing correctly

5)Display data on mirror

6)App to set schedule

Project Implementation Method

Following are the methods will used to perform each individual task:

Face detection

Python 3.0 software is used in order to detect human face. Open CV library will be used as it is the most popular library and widely used for the face detection and recognition techniques. Open CV uses machine learning algorithm in order to detect and recognize faces. As it is so complex and not easy to identify the faces accurately. As there are thousands of features that must be matched in order to successfully detect the human face. The Open CV algorithms basically break the task of identifying the face into thousands of smaller ones, which is easy to solve.

Face Recognition

Face Recognition will also detect human face as it was the first step of Face recognition. It will follows the principles that is used by Open CV. Once it will be detected successfully then it will do face recognition. For face recognition we must have a data set that will be created using video from which we take different angles and our data set will be created. Data set will created from video, in each second it take 4 clicks and new data set will be created for users. This will used for multiple users.

All data sets will be stored for different users. Every data has a frontal face that will converted into 128 bits. Whenever he detected a new face he will immediately will create a data set of different angles and will compare it different datasets of persons.

Open CV will applying deep learning in two key steps:

1. To apply face detection, which detects the presence and location of a face in an image, but does not identify it
2. To extract the 128-d feature vectors (called “embedding’s”) that quantify each face in an image

Posture Detection

Similarly, in order to detect the accuracy of posture we use OPEN POSE algorithm. Open pose also uses CNN technique, first to detect the human and then to find out the joints present in human body trained in algorithm. By using this technique, we get (x, y) coordinates for the further process. First we will collect data set of correct exercise video then we will train the algorithm on this data set. It will be done by tracking different body parts (used the COCO configuration with 18 key points).

After that we will test the algorithm which will compare the input with the data set.
We will have data sets of Pull up that will be compared with the joints of person that is performing exercise. It the comparison is good enough between stored data set and data set of the performing, else if the difference is spotted then it will tell user to perform it correctly.

Benefits of the Project

Virtual gym instructer not only benefit the Gym owners but also you can have it in your home. As it is the tensure of advance technology so it will be a advance equipement for Gyms.Following are the benefits that our project 

1. Cost efficient, no need to hire more trainers in the gym.
2. More accurate as compared to the other gym trainers. Once it is trained it can easily find errors in exercises.
3. Parallel operation, as it can be used for several people at the same time, every user will virtually have a dedicated trainer.
4. Customization, as it can be easily updated for multiple exercises.
5. You can have it at home or gym, where you have no access to physical trainer. It will be a perfect guide for user.
6. Convenience. It's always open regardless of the time of day or weather forecast.

Technical Details of Final Deliverable

• To detect user face using Open CV algorithm
• To display user gym routine
• To show demo video continuously

• To Monitor user exercise using Open pose algorithm
• To continuously display data of user
• To set schedule using app

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Health

Other Industries

Others

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT)

Sustainable Development Goals

Good Health and Well-Being for People

Required Resources

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