Merge branch 'main' of https://huggingface.co/spaces/EntrepreneurFirst/FitnessEquation

.gitattributes

@@ -34,3 +34,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 data/pose/squat.mp4 filter=lfs diff=lfs merge=lfs -text
+data/pose/squat_inference.mp4 filter=lfs diff=lfs merge=lfs -text

Modules/PoseEstimation/pose_estimator.py

@@ -0,0 +1,116 @@
+from ultralytics import YOLO
+import numpy as np
+
+id_joints_dict = {0: 'nose',
+        1: 'left_eye',
+        2: 'right_eye',
+        3: 'left_ear',
+        4: 'right_ear',
+        5: 'left_shoulder',
+        6: 'right_shoulder',
+        7: 'left_elbow',
+        8: 'right_elbow',
+        9: 'left_wrist',
+        10: 'right_wrist',
+        11: 'left_hip',
+        12: 'right_hip',
+        13: 'left_knee',
+        14: 'right_knee',
+        15: 'left_ankle',
+        16: 'right_ankle'}
+joints_id_dict = {v: k for k, v in id_joints_dict.items()}
+
+model = YOLO('yolov8n-pose.pt')
+
+def get_keypoints_from_keypoints(model, video_path):
+
+    keypoints = []
+    results = model(video_path, save=True, show_conf=False, show_boxes=False)
+    for frame in results:
+        keypoints.append(frame.keypoints.xy)
+
+    return keypoints
+
+keypoints = get_keypoints_from_keypoints(model, '../../data/pose/squat.mp4')
+
+def calculate_angle(a, b, c):
+
+    """
+    Calculates the angle between three joints.
+
+    Args:
+        a (tuple): coordinates of the first joint
+        b (tuple): coordinates of the second joint
+        c (tuple): coordinates of the third joint
+
+    Returns:
+        angle (float): angle between the three joints
+    """
+    
+    ba = np.array(a) - np.array(b)
+    bc = np.array(c) - np.array(b)
+
+    cosine_angle = np.dot(ba, bc) / (np.linalg.norm(ba) * np.linalg.norm(bc))
+    angle = np.arccos(cosine_angle)
+
+    return np.degrees(angle)
+
+def compute_left_knee_angle(pose):
+
+    """
+    Computes the knee angle.
+
+    Args:
+        pose (dict): pose dictionary
+
+    Returns:
+        knee_angle (float): knee angle
+    """
+
+    left_hip = pose[0][joints_id_dict['left_hip']]
+    left_knee = pose[0][joints_id_dict['left_knee']]
+    left_ankle = pose[0][joints_id_dict['left_ankle']]
+
+    knee_angle = calculate_angle(left_hip, left_knee, left_ankle)
+
+    return knee_angle
+
+def compute_right_knee_angle(pose):
+
+    """
+    Computes the knee angle.
+
+    Args:
+        pose (dict): pose dictionary
+
+    Returns:
+        knee_angle (float): knee angle
+    """
+
+    right_hip = pose[0][joints_id_dict['right_hip']]
+    right_knee = pose[0][joints_id_dict['right_knee']]
+    right_ankle = pose[0][joints_id_dict['right_ankle']]
+
+    knee_angle = calculate_angle(right_hip, right_knee, right_ankle)
+
+    return knee_angle
+
+def moving_average(data, window_size):
+
+    """
+    Computes the moving average of a list.
+
+    Args:
+        data (list): list of values
+        window_size (int): size of the window
+
+    Returns:
+        avg (list): list of moving average values
+    """
+
+    avg = []
+    for i in range(len(data) - window_size + 1):
+        avg.append(sum(data[i:i + window_size]) / window_size)
+
+    return avg
+

Modules/Speech2Text/transcribe.py

@@ -3,4 +3,4 @@ import whisper
 def transcribe(audio_path):
     model = whisper.load_model("base")
     result = model.transcribe(audio_path)
-    return result.text
+    return result['text']

app.py

@@ -1,32 +1,62 @@
 import streamlit as st
-from moviepy.editor import AudioClip
-import tempfile
-import os
 from st_audiorec import st_audiorec
+from Modules.Speech2Text.transcribe import transcribe
+import base64
 
 st.set_page_config(layout="wide", initial_sidebar_state="collapsed")
 # Create two columns
 col1, col2 = st.columns(2)
+video_uploaded = None
 
 # First column containers
 with col1:
     st.subheader("Audio Recorder")
+    recorded = False
+    temp_path = 'data/temp_audio/audio_file.wav'
     wav_audio_data = st_audiorec()
+    if wav_audio_data is not None:
+        with open(temp_path, 'wb') as f:
+            # Write the audio data to the file
+            f.write(wav_audio_data)
+        instruction = transcribe(temp_path)
+        print(instruction)
+        recorded = True
+
 
     st.subheader("LLM answering")
-    # Add your content here
+    if recorded:
+        if "messages" not in st.session_state:
+            st.session_state.messages = []
+        for message in st.session_state.messages:
+            with st.chat_message(message["role"]):
+                st.markdown(message["content"])
 
-    st.subheader("Movement Analysis")
-    # Add your content here
+        st.session_state.messages.append({"role": "user", "content": instruction})
+        with st.chat_message("user"):
+            st.markdown(instruction)
+
+        with st.chat_message("assistant"):
+            # Build answer from LLM
+            response = " to be DEFINED "# TO DO 
+        st.session_state.messages.append({"role": "assistant", "content": response})
 
+    st.subheader("Movement Analysis")
+        # TO DO 
 # Second column containers
 with col2:
     st.subheader("Sports Agenda")
-    # Add your content here
-
+        # TO DO
     st.subheader("Video Analysis")
-    video_path = "data/video/temp.mp4"
-    st.video(video_path)
+    ask_video = st.empty()
+    if video_uploaded is None:
+        video_uploaded = ask_video.file_uploader("Choose a video file", type=["mp4", "ogg", "webm"])
+    if video_uploaded:
+        ask_video.empty()
+        with st.spin("Processing video"):
+            pass # TO DO
+        _left, mid, _right = st.columns(3)
+        with mid:
+            st.video(video_uploaded)
 
     st.subheader("Graph Displayer")
-    # Add your content here
+        # TO DO

config.py

@@ -0,0 +1,3 @@
+# Pose estimation
+
+pose_mode_size = 'm'

data/pose/squat_inference.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eaf32617db792283036d8ee54e4a53e96e88efab820c4367ea28f9851ebb9811
+size 1784680