app.py

import os
import tensorflow as tf
import tensorflow_hub as hub
import numpy as np
import csv
import requests
import json
import logging
import scipy
from scipy.io import wavfile
from pydub import AudioSegment
import io
from io import BytesIO

# Load the model
model = hub.load('Audio_Multiple_v1')

def class_names_from_csv(class_map_csv_text):
    """Returns list of class names corresponding to score vector."""
    class_names = []
    with tf.io.gfile.GFile(class_map_csv_text) as csvfile:
        reader = csv.DictReader(csvfile)
        for row in reader:
            class_names.append(row['display_name'])
    return class_names

class_map_path = model.class_map_path().numpy()
class_names = class_names_from_csv(class_map_path)

def ensure_sample_rate(original_sample_rate, waveform, desired_sample_rate=16000):
    if original_sample_rate != desired_sample_rate:  # Resample waveform if required
        desired_length = int(round(float(len(waveform)) / original_sample_rate * desired_sample_rate))
        waveform = np.array(scipy.signal.resample(waveform, desired_length), dtype=np.float32)
    return desired_sample_rate, waveform

def convert_mp3_to_wav(mp3_data):
    audio = AudioSegment.from_file(io.BytesIO(mp3_data), format="mp3")
    wav_buffer = io.BytesIO()
    audio.export(wav_buffer, format='wav')
    wav_buffer.seek(0)
    return wav_buffer.getvalue()

def process_audio_file(file_data, url):
    try:
        sample_rate, wav_data = wavfile.read(BytesIO(file_data))      
        
        if wav_data.ndim > 1:                        
            wav_data = np.mean(wav_data, axis=1)
        sample_rate, wav_data = ensure_sample_rate(sample_rate, wav_data)

        waveform = wav_data / tf.int16.max

        scores, embeddings, spectrogram = model(waveform)      

        scores_np = scores.numpy()
        mean_scores = np.mean(scores, axis=0)

        inferred_class = class_names[mean_scores.argmax()]    
        
        confidence_threshold = 0.60                 
        confident_classes = set()

        exclusion_list = ['Mechanisms','Domestic animals, pets', 'Animal', 'Silence', 'Alarm', 'Wind chime', 'Water', 'Livestock, farm animals, working animals', 'Wild animals', 'Bleat', 'Siren', 'Computer keyboard', 'Toot', 'Shatter', 'Bird','Caw', 'Independent music', 'Tender music', 'Ocean', 'House music', 'Middle Eastern music', 'Swing music', 'Soul music', 'Shofar', 'Motor vehicle (road)', 'White noise','Pink noise', 'Cacophony', 'Sidetone', 'Static', 'Outside, rural or natural', 'Outside, urban or manmade', 'Inside, public space', 'Inside, large room or hall', 'Inside, small room', 'Sound effect']
        for frame_scores in scores_np:
            for i, score in enumerate(frame_scores):
                if score > confidence_threshold:
                    class_name = class_names[i]

                    if class_name =='Child speech, kid speaking':
                        class_name='Child speech'
                    elif class_name =='Vehicle horn, car horn, honking':
                        class_name='Vehicle horn'
                    elif class_name =='Railroad car, train wagon':
                        class_name='Train/wagon'
                    elif class_name=='Rail transport':
                        class_name='Train/wagon'

                    if class_name not in exclusion_list:
                        confident_classes.add(class_name)

        confident_classes = sorted(confident_classes)
        confident_classes_list = list(confident_classes)
        
        answer_dict = {'url': url, 'answer': confident_classes_list}
        return answer_dict
    
    except Exception as e:
        logging.error(f"Error processing {url}: {e}")
        return None

def get_audio_data(url):
    response = requests.get(url)
    response.raise_for_status()
    return response.content

# def send_results_to_api(data, result_url):
#     headers = {"Content-Type": "application/json"}
#     try:
#         response = requests.post(result_url, json=data, headers=headers)
#         response.raise_for_status()  # Raise error for non-200 responses
#         return response.json()  # Return any JSON response from the API
#     except requests.exceptions.HTTPError as http_err:
#         logging.error(f"HTTP error occurred: {http_err}")
#         return {"error": f"HTTP error occurred: {http_err}"}
#     except requests.exceptions.RequestException as req_err:
#         logging.error(f"Request error occurred: {req_err}")
#         return {"error": f"Request error occurred: {req_err}"}
#     except ValueError as val_err:
#         logging.error(f"Error decoding JSON response: {val_err}")
#         return {"error": f"Error decoding JSON response: {val_err}"}

def process_audio(params):
    try:
        params = json.loads(params)
    except json.JSONDecodeError as e:
        return {"error": f"Invalid JSON input: {e.msg} at line {e.lineno} column {e.colno}"}

    audio_files = params.get("urls", [])
    # api = params.get("api", "")
    # job_id = params.get("job_id", "")

    solutions = []
    for audio_url in audio_files:
        audio_data = get_audio_data(audio_url)

        if audio_url.endswith(".mp3"):            
            wav_data = convert_mp3_to_wav(audio_data)
            result = process_audio_file(wav_data, audio_url)

        elif audio_url.endswith(".wav"):           
            result = process_audio_file(audio_data, audio_url)
        
        if result:
            solutions.append(result)

    # result_url = f"{api}/{job_id}"
    # send_results_to_api(solutions, result_url)

    return json.dumps({"solutions": solutions})

import gradio as gr

inputt = gr.Textbox(label="Parameters (JSON format) Eg. {'urls':['file1.mp3','file2.wav']}")
outputs = gr.JSON()

application = gr.Interface(fn=process_audio, inputs=inputt, outputs=outputs, title="Audio Classification with API Integration")
application.launch()