Fitness training website

ABSTRACT: –

Nowadays, because of busy schedules, people have no time to go to the gym, and even if they manage to find a gym nearby, having a gym trainer besides all the time to correct postures while doing exercise is impossible unless people do not opt for a personal trainer. Even if they assist a personal trainer for them they would have to adjust their time accordingly and both of these methods are quite costly and not everyone can afford it, also during this global pandemic, people are stuck at home and have no access to go to the gym or can’t even take a risk of getting in contact with a personal trainer. Performing exercises let it be exercise or doing exercise proper body posture is important, if not performed properly it can lead to crucial problems such as poor joint alignment, increased shear forces on the spine, compression of discs and joints, less space for nerves to course through the body due to compression, reduced blood flow, etc. to prevent such injuries and pains, and to track gym exercises repetitions we came up with a system called Fitness Freaks. Fitness Freaks is an AI Fitness tracker. It tracks users’ body movements using human pose estimation. This in turn keeps track of repetitions of gym exercises and detection of wrong body posture while doing exercise.

If we search online for workout applications, we get many results with multiple functionalities. These applications provide many workout programs which help us to perform it on our own. But somewhere those programs cannot improve the user’s posture and accuracy. Those applications aim to provide workouts only, but sometimes following these workouts in the wrong way may lead to short-term or permanent injuries. To avoid such a problem, we proposed a system for workout analysis using the pose classification technique. The objective is to develop an application that can assist people in performing various exercises without getting injured. An application with the pre-trained workout set with the help of pose estimation technique. We proposed to use the workout -pose pose estimation module developed by Media   Pipe. This neural network provides 33 body-points which are more than enough to capture the movements of the user. The pose estimation model is generally used to classify the different movements. We are using such technology with some advancement to provide accuracy of the user’s workout which will provide state of art results. The application will not just provide workouts to users but also it will monitor the real-time movements of users and also provide accuracy to users.

SYSTEM:-

• User Creation: Users create a profile that includes their health goals, current health levels, dietary preferences, and general health information.

• Creation of personalized fitness plan: The system uses artificial intelligence algorithms such as support learning and genetic algorithm to create a fitness plan for each user. Fitness programs include exercise, meal planning, and lifestyle advice.

• Real-Time Feedback and Monitoring: This system uses data from the user’s fitness and other IoT devices to provide real-time feedback and monitor the user’s progress. The system adjusts the physical program as needed based on the user’s progress and feedback.
• AI-powered virtual assistant: The system includes an AI-powered virtual assistant that can answer user questions, provide support and encouragement, and make personalized recommendations.

• Motion Recognition and Correction: This system uses computer vision and machine learning algorithms to recognize user’s movements and provide real-time feedback on motion patterns and strategies. The system can also suggest corrections and adjustments to prevent injury and improve results.

• Interaction and Gamification: The system includes interactions and games such as leaderboards, contests and virtual prizes to encourage and engage users.

• Data Analysis and Reporting: The system provides data analysis and reporting capabilities such as personal achievement data and insights to help users and coaches track their progress and adjust their fitness programs as needed.

PROPOSED SYSTEM:-

The proposed system is limited to 6 yoga postures where there are more than 80 yoga postures in total. The suggested data set can be extended by adding the desired yoga pose key points. The technology can also be used to perform real-time predictions and self-training on a mobile device. There are several examples of real-life applications where a single individual location assessment will not suffice; for example, position estimation in a crowded environment will need to detect and recognize the position of each participant. It is quite challenging to include many positions and get models in many positions (classification of many positions). The Keras position estimate affects model performance; Steps should be taken to get key points when body parts are overlapping or missing to get better results. This method for extracting angles as features can be used for other applications such as activity detection and sports activity tracking.

first, all the joints involved in the exercise are identified, then the number pointing to the joints is identified using the coconut human position estimation model, then the professional trainer’s video is used to detect the ideal exercise movements, and for the next exercise, the position estimation is used to track the professional trainer’s joint movements, they are detected angles between the joints and a certain threshold value is maintained to neglect the disproportionality caused by different body types and sizes depending on gender and age when exercising as well as detection of exercises under one roof for biceps curl exercises the desired distance angle and limb movement are taught to the system using trainer videos and for Warrior ii exercises the position of the professional exercise instructor is recorded inside the system, once the rules are captured the video of the user is fed into the system and the moments are detected using computer vision i.e. position estimation using o The pencv rules are checked based on the exercise performed by tensorflow errors are detected and repetitions are tracked for the biceps curl exercise.

• Human position estimation
  • Pose estimation is a machine learning task that estimates the position of a person from an image or video by estimating the spatial location of specific body parts (key points). Position estimation is a computer vision technique to track the movement of a person or object. This is usually done by finding the location of the key points for the given objects. Based on these key points, we can compare different movements and positions and gain insights.

• Position estimation with deep learning
  • With the rapid development of deep learning solutions in recent years, it has been shown that deep learning outperforms classical computer vision methods in various tasks, including image segmentation or object detection. Therefore, deep learning techniques have brought significant progress and performance enhancement in position estimation tasks. There are many deep learning estimation approaches available, e.g., openpose, movenet, deeppose, posnet, bodynet[6] etc. In our scenario, we use training pose because it is the latest model developed by Google and works smoothly on lightweight devices like browsers or mobile devices.

• Estimated position for one person
  • The single position estimation algorithm is the simpler and faster of the two. It is an ideal use case when there is only one person in the center of the input image or video. The downside is that if there are multiple people in the image, keypoints from both people are likely to be estimated as part of the same single pose – meaning that, for example, Person 1’s left arm and Person 2’s right knee might be associated by the algorithm as belonging to the same pose.

MODULES:-

• user profile module: this module enables users to create a profile that includes their fitness goals current fitness level dietary preferences and any health conditions the module can use machine learning algorithms to analyze user input and suggest personalized fitness plans based on their profile.
• workout plan module: this module generates personalized workout plans for users based on their profile using ai algorithms such as reinforcement learning and genetic algorithms the workout plan includes exercises sets reps rest periods and progression schemes tailored to the users fitness level and goals
• meal plan module: this module generates personalized meal plans for users based on their profile using ai algorithms such as natural language processing and rule-based systems the meal plan includes recipes portion sizes and nutrient targets tailored to the users dietary preferences and goals
• exercise recognition module: this module uses computer vision and machine learning algorithms to recognize users exercises and provide real-time feedback on their form and technique the module can also suggest corrections and modifications to prevent injury and optimize results
• virtual assistant module: this module includes an ai-powered virtual assistant that can answer users questions provide encouragement and motivation and make personalized recommendations the virtual assistant can use natural language processing and sentiment analysis to understand user input and respond in a human-like manner
• data analytics module: this module provides data analytics and reporting capabilities such as personalized progress reports and insights to help users and trainers track their progress and adjust their fitness plans as needed the module can use data visualization and machine learning algorithms to identify trends and patterns in user data
• social interaction module :this module includes social interaction and gamification features such as leaderboards challenges and virtual rewards to motivate and engage users the module can use reinforcement learning and game theory to design effective incentives and rewards for users.
• integration module :this module integrates with third-party services such as fitness equipment meal delivery services and wellness apps to provide users with a seamless and integrated fitness experience the module can use application programming interfaces apis and webhooks to connect with external services

these modules can work together to provide users with a personalized and effective fitness experience that adapts to their needs and preferences in real-time the ai-powered virtual assistant exercise recognition and correction and social interaction and gamification features can help users stay motivated and engaged while the data analytics and reporting capabilities can help users and trainers track progress and optimize their fitness plans

APPLICATION:-

• Personalized Workout Plan: The app creates workout plans based on your fitness level, goals and time. The app uses machine learning algorithms to adjust the plan as you progress and suggests keeping track of your goals and interests.

• Motion Recognition: The app uses computer vision technology to recognize your movements and provide real-time feedback on your form and technique. The app can suggest corrections and adjustments to help you improve your technique and prevent injury.

• Meal Planning: The app creates personalized meal plans based on your meal preferences, goals and diet.
• The app uses natural language processing and rule-based techniques to suggest recipes and sizes based on your needs.

• Virtual Assistant: This app includes a virtual assistant that can answer your questions, provide motivation and support, and make personalized recommendations. Virtual assistants use language processing and sentiment analysis to understand your input and respond in a humane way.

• Analytics: The app includes data analysis and reporting features that allow you to track your progress and adjust your fitness program as needed. The app provides personal success reports and insights to help you reach your goals.
• Interactive: The app includes social and game-like leaderboards, contests, and virtual prizes to encourage and engage you. The app uses motivational learning and game theory to create positive incentives and rewards.

• Integration with third-party services: The app integrates with third-party services such as fitness apps, nutrition services, and health apps to provide healthy nutrition services for you.

Fitness AI is the best app for anyone who wants to achieve their fitness goals with the help of smart technology. Whether you’re a beginner or a seasoned athlete, Fitness AI can help you improve your workouts and nutrition and stay on top of things.

HARDWARE AND SOFTWARE REQUIREMENTS:-

HARDWARE:-

·        Processor: Intel Core i3 or more.

·        RAM: 4GB or more.

·        Hard disk: 250 GB or more.

SOFTWARE:-

• Operating System: Windows 10, 7, 8.
• python
• anaconda
• Spyder, Jupyter notebook, Flask.
• MYSQL