Upload remove_silence_from_samples.ipynb

remove_silence_from_samples.ipynb

@@ -0,0 +1,66 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from pydub import AudioSegment\n",
+    "import os, librosa, soundfile as sf\n",
+    "from pathlib import Path\n",
+    "from tqdm import tqdm\n",
+    "\n",
+    "top_db=20 #in decibels\n",
+    "threshold=-20.0 #in decibels\n",
+    "\n",
+    "def milliseconds_until_sound(sound, threshold, chunk_size=10):\n",
+    "    trim_ms = 0  # ms\n",
+    "    assert chunk_size > 0  # to avoid infinite loop\n",
+    "    while sound[trim_ms:trim_ms+chunk_size].dBFS < threshold and trim_ms < len(sound):\n",
+    "        trim_ms += chunk_size\n",
+    "    return trim_ms\n",
+    "\n",
+    "def remove_silence(input_file, output_file, top_db=top_db):#, target_sr=None):\n",
+    "    sr = 16000\n",
+    "    y, sr = librosa.load(input_file)\n",
+    "    # if target_sr and target_sr != sr:\n",
+    "    #     y = librosa.resample(y, orig_sr=sr, target_sr=target_sr)\n",
+    "    #     sr = target_sr\n",
+    "    intervals = librosa.effects.split(y, top_db=top_db)\n",
+    "    y_trimmed = []\n",
+    "    for start, end in intervals:\n",
+    "        y_trimmed.extend(y[start:end])\n",
+    "    sf.write(output_file, y_trimmed, sr)\n",
+    "\n",
+    "def process_directory(input_dir, output_dir, top_db=top_db):#, target_sr=None):\n",
+    "    if not os.path.exists(output_dir):\n",
+    "        os.makedirs(output_dir)\n",
+    "\n",
+    "    for filename in os.listdir(input_dir):\n",
+    "        if filename.endswith(\".mp3\"):\n",
+    "            input_file = os.path.join(input_dir, filename)\n",
+    "            output_file = os.path.join(output_dir, filename)\n",
+    "            audio = AudioSegment.from_mp3(input_file)\n",
+    "            # if audio.dBFS < -30:\n",
+    "            start_trim = milliseconds_until_sound(audio)\n",
+    "            if start_trim < len(audio):\n",
+    "                remove_silence(input_file, output_file, top_db)#, target_sr)\n",
+    "            # else:\n",
+    "            #     (print(\"file removed\"))\n",
+    "\n",
+    "if __name__ == \"__main__\":\n",
+    "    input_dir = folder_path\n",
+    "    output_dir = folder_path + \"_trim/\"\n",
+    "    process_directory(input_dir, output_dir)#, target_sr=16000)"
+   ]
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}