Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| import json | |
| from streamlit_lottie import st_lottie | |
| # Load Lottie animation from a JSON file | |
| def load_lottiefile(filepath: str): | |
| with open(filepath, "r") as f: | |
| return json.load(f) | |
| st.set_page_config(layout="wide") | |
| st.markdown("<h1 style='text-align: center;'>Workout Easz</h1>", unsafe_allow_html=True) | |
| lottie_animation = load_lottiefile("ltt.json") | |
| st_lottie(lottie_animation, height=300, key="header_animation") | |
| # Disclaimer | |
| st.markdown(""" | |
| <div style="text-align: center; color: #FF0000;"> | |
| <strong>Disclaimer:</strong> | |
| The performance of pose detection depends on the model's accuracy. As of now, our application uses a model with approximately 50% pose detection accuracy. Results may vary based on the quality of the video input and the model's limitations. | |
| </div> | |
| """, unsafe_allow_html=True) | |
| # Introduction and Key Points | |
| with st.expander("Introduction to MediaPipe"): | |
| st.header("Introduction to MediaPipe") | |
| st.write(""" | |
| MediaPipe is an open-source framework developed by Google for building multimodal machine learning pipelines. Its Pose solution is designed for high-fidelity body pose tracking, detecting 33 key body landmarks in real-time[1][2]. | |
| """) | |
| with st.expander("Extraction of 33 Key Points"): | |
| st.header("Extraction of 33 Key Points") | |
| st.write(""" | |
| MediaPipe Pose detects 33 key body landmarks, each with x, y, z coordinates, and a visibility score. These landmarks represent various parts of the body, such as the nose, eyes, ears, shoulders, elbows, wrists, hips, knees, ankles, and feet[1][2]. | |
| """) | |
| st.image("pose_landmarks_index.png", caption="MediaPipe Pose Landmarks", width=400) | |
| # Squat Analysis | |
| squat_expander = st.expander("Squat Form Analysis") | |
| with squat_expander: | |
| st.header("Squat Form Analysis") | |
| # Key Landmarks for Squat Analysis | |
| st.subheader("1. Key Landmarks for Squat Analysis") | |
| st.write(""" | |
| For squat analysis, we focus on the following key landmarks: | |
| - Shoulder (left and right) | |
| - Hip (left and right) | |
| - Knee (left and right) | |
| - Ankle (left and right) | |
| These landmarks are essential for calculating the joint angles critical for analyzing squat form[3]. | |
| """) | |
| # Calculation of Joint Angles for Squat | |
| st.subheader("2. Calculation of Joint Angles for Squat") | |
| # Knee Angle | |
| st.write("**a. Knee Angle**") | |
| st.write(""" | |
| The knee angle is formed by the hip, knee, and ankle. It is calculated using the arctangent function: | |
| """) | |
| st.latex(r"\theta_{\text{knee}} = \left| \frac{180}{\pi} \times \left( \arctan2(y_{\text{ankle}} - y_{\text{knee}}, x_{\text{ankle}} - x_{\text{knee}}) - \arctan2(y_{\text{hip}} - y_{\text{knee}}, x_{\text{hip}} - x_{\text{knee}}) \right) \right|") | |
| # Hip Angle | |
| st.write("**b. Hip Angle**") | |
| st.write(""" | |
| The hip angle is formed by the shoulder, hip, and knee. It is calculated similarly: | |
| """) | |
| st.latex(r"\theta_{\text{hip}} = \left| \frac{180}{\pi} \times \left( \arctan2(y_{\text{knee}} - y_{\text{hip}}, x_{\text{knee}} - x_{\text{hip}}) - \arctan2(y_{\text{shoulder}} - y_{\text{hip}}, x_{\text{shoulder}} - x_{\text{hip}}) \right) \right|") | |
| # Back Angle | |
| st.write("**c. Back Angle**") | |
| st.write(""" | |
| The back angle is formed by the shoulder, hip, and ankle: | |
| """) | |
| st.latex(r"\theta_{\text{back}} = \left| \frac{180}{\pi} \times \left( \arctan2(y_{\text{ankle}} - y_{\text{hip}}, x_{\text{ankle}} - x_{\text{hip}}) - \arctan2(y_{\text{shoulder}} - y_{\text{hip}}, x_{\text{shoulder}} - x_{\text{hip}}) \right) \right|") | |
| # Form Issue Detection for Squat | |
| st.subheader("3. Form Issue Detection for Squat") | |
| st.write("**Ideal Angle Ranges:**") | |
| st.write(""" | |
| - Knee Angle: 90 to 110 degrees | |
| - Hip Angle: 80 to 100 degrees | |
| - Back Angle: 70 to 90 degrees | |
| """) | |
| st.write("**Detection of Deviations:**") | |
| st.write(""" | |
| - Knees Bending Too Much: Knee angle < 90 degrees | |
| - Hips Bending Too Much: Hip angle < 80 degrees | |
| - Back Leaning Too Much: Back angle < 70 degrees | |
| """) | |
| # Providing Suggestions for Squat | |
| st.subheader("4. Providing Suggestions for Squat") | |
| st.write(""" | |
| Based on the detected form issues, specific suggestions are provided: | |
| 1. Knees Bending Too Much: | |
| - Suggestion: "Watch your knee bend. Ensure your knees do not bend excessively and maintain proper alignment with your toes." | |
| 2. Hips Bending Too Much: | |
| - Suggestion: "Keep your hips higher. Avoid excessive hip bending by maintaining a more upright posture." | |
| 3. Back Leaning Too Much: | |
| - Suggestion: "Maintain a straighter back. Focus on keeping your back straight and avoid leaning forward excessively." | |
| """) | |
| # Push-up Analysis | |
| pushup_expander = st.expander("Push-up Form Analysis") | |
| with pushup_expander: | |
| st.header("Push-up Form Analysis") | |
| # Key Landmarks for Push-up Analysis | |
| st.subheader("1. Key Landmarks for Push-up Analysis") | |
| st.write(""" | |
| For push-up analysis, we focus on the following key landmarks: | |
| - Shoulder (left and right) | |
| - Elbow (left and right) | |
| - Wrist (left and right) | |
| - Hip (left and right) | |
| - Ankle (left and right) | |
| These landmarks are crucial for calculating the joint angles and body alignment in push-ups[3]. | |
| """) | |
| # Calculation of Joint Angles for Push-up | |
| st.subheader("2. Calculation of Joint Angles for Push-up") | |
| # Elbow Angle | |
| st.write("**a. Elbow Angle**") | |
| st.write(""" | |
| The elbow angle is formed by the shoulder, elbow, and wrist. It is calculated using: | |
| """) | |
| st.latex(r"\theta_{\text{elbow}} = \left| \frac{180}{\pi} \times \left( \arctan2(y_{\text{wrist}} - y_{\text{elbow}}, x_{\text{wrist}} - x_{\text{elbow}}) - \arctan2(y_{\text{shoulder}} - y_{\text{elbow}}, x_{\text{shoulder}} - x_{\text{elbow}}) \right) \right|") | |
| # Body Alignment Angle | |
| st.write("**b. Body Alignment Angle**") | |
| st.write(""" | |
| The body alignment angle is formed by the shoulder, hip, and ankle. It is calculated as: | |
| """) | |
| st.latex(r"\theta_{\text{alignment}} = \left| \frac{180}{\pi} \times \left( \arctan2(y_{\text{ankle}} - y_{\text{hip}}, x_{\text{ankle}} - x_{\text{hip}}) - \arctan2(y_{\text{shoulder}} - y_{\text{hip}}, x_{\text{shoulder}} - x_{\text{hip}}) \right) \right|") | |
| # Form Issue Detection for Push-up | |
| st.subheader("3. Form Issue Detection for Push-up") | |
| st.write("**Ideal Angle Ranges:**") | |
| st.write(""" | |
| - Elbow Angle: 90 to 110 degrees (at the bottom of the push-up) | |
| - Body Alignment Angle: 170 to 180 degrees (straight body) | |
| """) | |
| st.write("**Detection of Deviations:**") | |
| st.write(""" | |
| - Incomplete Push-up: Elbow angle > 110 degrees at the bottom | |
| - Elbow Flaring: Elbow angle < 90 degrees | |
| - Sagging Hips: Body alignment angle < 170 degrees | |
| """) | |
| # Providing Suggestions for Push-up | |
| st.subheader("4. Providing Suggestions for Push-up") | |
| st.write(""" | |
| Based on the detected form issues, specific suggestions are provided: | |
| 1. Incomplete Push-up: | |
| - Suggestion: "Lower your body further. Aim to bring your chest closer to the ground for a full range of motion." | |
| 2. Elbow Flaring: | |
| - Suggestion: "Keep your elbows closer to your body. Tuck them in to engage your chest and triceps more effectively." | |
| 3. Sagging Hips: | |
| - Suggestion: "Maintain a straight body line. Engage your core to keep your hips aligned with your shoulders and ankles." | |
| """) | |
| # Conclusion | |
| with st.expander("Conclusion",expanded=False): | |
| st.write(""" | |
| By leveraging MediaPipe's pose estimation capabilities, this application can effectively analyze both squat and push-up forms. It calculates key joint angles and body alignments, compares them to ideal ranges, and provides specific suggestions for improvement. This approach enables users to receive real-time feedback on their exercise technique, helping them to perform exercises correctly and reduce the risk of injury[1][2][3]. | |
| """) | |
| # Version Updates | |
| with st.expander("Version Updates", expanded=False): | |
| st.header("Version Updates") | |
| st.subheader("Current Version") | |
| st.write(""" | |
| **Current Update:** As of today, our application now includes both squat and push-up form analysis. | |
| """) | |
| st.subheader("Previous Versions") | |
| st.write(""" | |
| **Initial Release:** Squat workout analysis, released on July 19th. | |
| """) | |
| st.write(""" | |
| Stay tuned for more updates as we continue to enhance our application's features and performance. | |
| """) | |
| # References | |
| with st.expander("References"): | |
| st.header("References") | |
| st.write(""" | |
| [1] [MediaPipe Pose Documentation](https://github.com/google-ai-edge/mediapipe/blob/master/docs/solutions/pose.md) | |
| [2] [Google AI Edge - MediaPipe Pose Landmarker](https://ai.google.dev/edge/mediapipe/solutions/vision/pose_landmarker) | |
| [3] [Real-time Human Pose Estimation using MediaPipe](https://sigmoidal.ai/en/real-time-human-pose-estimation-using-mediapipe/) | |
| [4] [Kaggle Dataset for Push-Up Analysis](https://www.kaggle.com/datasets/mohamadashrafsalama/pushup) | |
| """) | |