Update app.py

app.py

@@ -8,7 +8,7 @@ from tensorflow.keras.models import load_model
 import tensorflow_hub as hub
 import time
 import tempfile
-import streamlit.components.v1 as components
+import base64
 from io import BytesIO
 
 # Attempt to set GPU memory growth
@@ -65,27 +65,46 @@ def load_autism_model():
 
 model = load_autism_model()
 
-def extract_features(audio_bytes):
+def extract_features(path):
     sample_rate = 16000
-    with tempfile.NamedTemporaryFile(delete=True) as temp_audio_file:
-        temp_audio_file.write(audio_bytes)
-        temp_audio_file.flush()
-        array, fs = torchaudio.load(temp_audio_file.name)
+    array, fs = torchaudio.load(path)
 
     array = np.array(array)
     if array.shape[0] > 1:
         array = np.mean(array, axis=0, keepdims=True)
 
-    array = array[:, :sample_rate * 10]  # Truncate to 10 seconds
+    # Truncate the audio to 10 seconds for reducing memory usage
+    array = array[:, :sample_rate * 10]
+
     embeddings = m(array)['embedding']
     embeddings.shape.assert_is_compatible_with([None, 1024])
     embeddings = np.squeeze(np.array(embeddings), axis=0)
 
     return embeddings
 
-st.markdown('<span style="color:black; font-size: 48px; font-weight: bold;">Neu</span> <span style="color:black; font-size: 48px; font-weight: bold;">RO:</span> <span style="color:black; font-size: 48px; font-weight: bold;">An Application for Code-Switched Autism Detection in Children</span>', unsafe_allow_html=True)
+def save_temp_audio(base64_audio, filename="temp_audio.wav"):
+    audio_data = base64.b64decode(base64_audio)
+    temp_audio_path = tempfile.NamedTemporaryFile(delete=False, suffix=".wav")
+    with open(temp_audio_path.name, "wb") as f:
+        f.write(audio_data)
+    return temp_audio_path.name
 
-option = st.radio("**Choose an option:**", ["Upload an audio file", "Record audio"])
+def handle_audio_upload():
+    json_data = st.experimental_get_query_params().get("upload-audio")
+    if json_data:
+        base64_audio = json_data['audio_data']
+        temp_audio_path = save_temp_audio(base64_audio)
+        
+        # Process the uploaded audio file
+        command = f'ffmpeg -i {temp_audio_path} -acodec pcm_s16le -ar 16000 -ac 1 ./recorded_audio2.wav'
+        result = subprocess.run(command, shell=True, capture_output=True, text=True)
+        if result.returncode != 0:
+            st.error(f"Error running ffmpeg: {result.stderr}")
+        else:
+            features = extract_features("./recorded_audio2.wav")
+            run_prediction(features)
+            os.remove("./recorded_audio2.wav")
+            os.remove(temp_audio_path)
 
 def run_prediction(features):
     try:
@@ -158,13 +177,21 @@ def run_prediction(features):
                     unsafe_allow_html=True
                 )
 
+st.markdown('<span style="color:black; font-size: 48px; font-weight: bold;">Neu</span> <span style="color:black; font-size: 48px; font-weight: bold;">RO:</span> <span style="color:black; font-size: 48px; font-weight: bold;">An Application for Code-Switched Autism Detection in Children</span>', unsafe_allow_html=True)
+
+option = st.radio("**Choose an option:**", ["Upload an audio file", "Record audio"])
+
 if option == "Upload an audio file":
     uploaded_file = st.file_uploader("Upload an audio file (.wav)", type=["wav"])
     if uploaded_file is not None:
-        start_time = time.time()
+        start_time = time.time()  # Record start time
         with st.spinner('Extracting features...'):
-            audio_bytes = uploaded_file.read()
-            features = extract_features(audio_bytes)
+            # Process the uploaded file
+            temp_audio_path = os.path.join(".", "temp_audio.wav")
+            with open(temp_audio_path, "wb") as f:
+                f.write(uploaded_file.getbuffer())
+            features = extract_features(temp_audio_path)
+            os.remove(temp_audio_path)
             run_prediction(features)
         elapsed_time = round(time.time() - start_time, 2)
         st.write(f"Elapsed Time: {elapsed_time} seconds")
@@ -254,15 +281,24 @@ else:  # Option is "Record audio"
                     };
                     recorder.onstop = () => {
                         const audioBlob = new Blob(audioChunks, { type: 'audio/wav' });
-                        const audioUrl = URL.createObjectURL(audioBlob);
-                        const a = document.createElement('a');
-                        a.href = audioUrl;
-                        a.download = 'recorded_audio.wav';
-                        document.body.appendChild(a);
-                        a.click();
-                        // Instead of downloading, pass the Blob data to the backend
-                        window.parent.postMessage(audioBlob, "*");
-
+                        const reader = new FileReader();
+                        reader.readAsDataURL(audioBlob);
+                        reader.onloadend = () => {
+                            const base64String = reader.result.split(',')[1];
+                            fetch('/upload-audio', {
+                                method: 'POST',
+                                headers: {
+                                    'Content-Type': 'application/json',
+                                },
+                                body: JSON.stringify({ audio_data: base64String }),
+                            }).then(response => {
+                                if (response.ok) {
+                                    console.log('Audio uploaded successfully.');
+                                } else {
+                                    console.error('Audio upload failed.');
+                                }
+                            });
+                        };
                         // Reset
                         audioChunks = [];
                         clearInterval(timerInterval);
@@ -292,12 +328,4 @@ else:  # Option is "Record audio"
     '''
     st.components.v1.html(audio_recorder_html, height=600)
 
-    if st.button("Click to Predict"):
-        try:
-            # Replace this with the actual method to receive audio bytes from frontend
-            audio_bytes = st.session_state.get('recorded_audio_bytes')
-            if audio_bytes:
-                features = extract_features(audio_bytes)
-                run_prediction(features)
-        except Exception as e:
-            st.error(f"An error occurred: {e}")
+    handle_audio_upload()