svvhsq

app.py

@@ -3,7 +3,6 @@ import tempfile
 import warnings
 from flask import Flask, request, jsonify
 from inference_service import (
-    load_hear_model,
     load_classifier,
     preprocess_audio,
     generate_embeddings,
@@ -21,8 +20,7 @@ warnings.filterwarnings("ignore", module="librosa")
 
 app = Flask(__name__)
 
-# Load models globally
-hear_infer_fn = load_hear_model()
+# Load classifier globally
 classifier_model = load_classifier()
 
 @app.route('/predict_pneumonia', methods=['POST'])
@@ -39,8 +37,8 @@ def predict_pneumonia_endpoint():
     audio_file.save(temp_audio_path)
 
     try:
-        if hear_infer_fn is None or classifier_model is None:
-            return jsonify({"error": "Models not loaded"}), 500
+        if classifier_model is None:
+            return jsonify({"error": "Classifier not loaded"}), 500
 
         audio_clips = preprocess_audio(
             temp_audio_path, SAMPLE_RATE, CLIP_DURATION, CLIP_OVERLAP_PERCENT,
@@ -50,12 +48,18 @@ def predict_pneumonia_endpoint():
         if audio_clips.size == 0:
             return jsonify({"result": "No valid audio clips", "risk_score": None}), 200
 
-        embeddings = generate_embeddings(audio_clips, hear_infer_fn)
+        # Generate embeddings using OpenL3
+        embeddings = generate_embeddings(audio_clips)
 
         if embeddings.size == 0:
             return jsonify({"result": "No embeddings generated", "risk_score": None}), 200
 
+        # Predict pneumonia risk
         clip_predictions, clip_probabilities = predict_pneumonia(embeddings, classifier_model)
+        
+        if clip_predictions is None or clip_probabilities is None:
+            return jsonify({"result": "Prediction failed", "risk_score": None}), 200
+
         final_prediction, risk_score = aggregate_predictions(clip_predictions, clip_probabilities)
 
         return jsonify({