{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"pygments_lexer":"ipython3","nbconvert_exporter":"python","version":"3.6.4","file_extension":".py","codemirror_mode":{"name":"ipython","version":3},"name":"python","mimetype":"text/x-python"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"import tensorflow as tf\nimport tensorflow_hub as hub\nimport tensorflow_io as tfio\n\nimport pandas as pd\nimport numpy as np\nimport librosa\nimport glob\n\nimport csv\nimport io\n\nfrom IPython.display import Audio","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:22.729102Z","iopub.execute_input":"2023-05-26T13:42:22.730040Z","iopub.status.idle":"2023-05-26T13:42:31.400735Z","shell.execute_reply.started":"2023-05-26T13:42:22.729989Z","shell.execute_reply":"2023-05-26T13:42:31.399562Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"model = hub.load('https://kaggle.com/models/google/bird-vocalization-classifier/frameworks/tensorFlow2/variations/bird-vocalization-classifier/versions/2')\nlabels_path = hub.resolve('https://kaggle.com/models/google/bird-vocalization-classifier/frameworks/tensorFlow2/variations/bird-vocalization-classifier/versions/2') + \"/assets/label.csv\"","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:31.402857Z","iopub.execute_input":"2023-05-26T13:42:31.403484Z","iopub.status.idle":"2023-05-26T13:42:37.439711Z","shell.execute_reply.started":"2023-05-26T13:42:31.403449Z","shell.execute_reply":"2023-05-26T13:42:37.438676Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"# Find the name of the class with the top score when mean-aggregated across frames.\ndef class_names_from_csv(class_map_csv_text):\n    \"\"\"Returns list of class names corresponding to score vector.\"\"\"\n    with open(labels_path) as csv_file:\n        csv_reader = csv.reader(csv_file, delimiter=',')\n        class_names = [mid for mid, desc in csv_reader]\n        return class_names[1:]\n\n\nclasses = class_names_from_csv(labels_path)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.440949Z","iopub.execute_input":"2023-05-26T13:42:37.441224Z","iopub.status.idle":"2023-05-26T13:42:37.454009Z","shell.execute_reply.started":"2023-05-26T13:42:37.441201Z","shell.execute_reply":"2023-05-26T13:42:37.453228Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"train_metadata = pd.read_csv(\"/kaggle/input/birdclef-2023/train_metadata.csv\")\ntrain_metadata.head()\ncompetition_classes = sorted(train_metadata.primary_label.unique())\n\nforced_defaults = 0\ncompetition_class_map = []\nfor c in competition_classes:\n    try:\n        i = classes.index(c)\n        competition_class_map.append(i)\n    except:\n        competition_class_map.append(0)\n        forced_defaults += 1\n        \n#count of classes we cant predict\nforced_defaults","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.456508Z","iopub.execute_input":"2023-05-26T13:42:37.457518Z","iopub.status.idle":"2023-05-26T13:42:37.634217Z","shell.execute_reply.started":"2023-05-26T13:42:37.457456Z","shell.execute_reply":"2023-05-26T13:42:37.633019Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#load the audio provided and break it up into the five-second samples with a sample rate of 32,000 \n\ndef frame_audio(\n      audio_array: np.ndarray,\n      window_size_s: float = 5.0,\n      hop_size_s: float = 5.0,\n      sample_rate = 32000,\n      ) -> np.ndarray:\n    \n    \"\"\"Helper function for framing audio for inference.\"\"\"\n    \"\"\" using tf.signal \"\"\"\n    if window_size_s is None or window_size_s < 0:\n        return audio_array[np.newaxis, :]\n    frame_length = int(window_size_s * sample_rate)\n    hop_length = int(hop_size_s * sample_rate)\n    framed_audio = tf.signal.frame(audio_array, frame_length, hop_length, pad_end=True)\n    return framed_audio\n\ndef ensure_sample_rate(waveform, original_sample_rate,\n                       desired_sample_rate=32000):\n    \"\"\"Resample waveform if required.\"\"\"\n    if original_sample_rate != desired_sample_rate:\n        waveform = tfio.audio.resample(waveform, original_sample_rate, desired_sample_rate)\n    return desired_sample_rate, waveform","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.635517Z","iopub.execute_input":"2023-05-26T13:42:37.635839Z","iopub.status.idle":"2023-05-26T13:42:37.642981Z","shell.execute_reply.started":"2023-05-26T13:42:37.635812Z","shell.execute_reply":"2023-05-26T13:42:37.642287Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"#submission\ndef predict_for_sample(filename, sample_submission, frame_limit_secs=None):\n    file_id = filename.split(\".ogg\")[0].split(\"/\")[-1]\n    \n    audio, sample_rate = librosa.load(filename)\n    sample_rate, wav_data = ensure_sample_rate(audio, sample_rate)\n    \n    fixed_tm = frame_audio(wav_data)\n    \n    frame = 5\n    all_logits, all_embeddings = model.infer_tf(fixed_tm[:1])\n    for window in fixed_tm[1:]:\n        if frame_limit_secs and frame > frame_limit_secs:\n            continue\n        \n        logits, embeddings = model.infer_tf(window[np.newaxis, :])\n        all_logits = np.concatenate([all_logits, logits], axis=0)\n        frame += 5\n    \n    frame = 5\n    all_probabilities = []\n    for frame_logits in all_logits:\n        probabilities = tf.nn.softmax(frame_logits).numpy()\n        \n        ## set the appropriate row in the sample submission\n        sample_submission.loc[sample_submission.row_id == file_id + \"_\" + str(frame), competition_classes] = probabilities[competition_class_map]\n        frame += 5","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.643841Z","iopub.execute_input":"2023-05-26T13:42:37.644579Z","iopub.status.idle":"2023-05-26T13:42:37.663151Z","shell.execute_reply.started":"2023-05-26T13:42:37.644554Z","shell.execute_reply":"2023-05-26T13:42:37.662176Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"test_samples = list(glob.glob(\"/kaggle/input/birdclef-2023/test_soundscapes/*.ogg\"))\ntest_samples","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.664434Z","iopub.execute_input":"2023-05-26T13:42:37.664714Z","iopub.status.idle":"2023-05-26T13:42:37.682967Z","shell.execute_reply.started":"2023-05-26T13:42:37.664690Z","shell.execute_reply":"2023-05-26T13:42:37.681877Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"pred = pd.DataFrame()\npred = {'row_id': []}\ncompetition_classes.sort()\n\nfor bird in competition_classes:\n    pred[bird]=[]\npred = pd.DataFrame(pred)\npred","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.684091Z","iopub.execute_input":"2023-05-26T13:42:37.684347Z","iopub.status.idle":"2023-05-26T13:42:37.718262Z","shell.execute_reply.started":"2023-05-26T13:42:37.684321Z","shell.execute_reply":"2023-05-26T13:42:37.717258Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sample_sub = pd.read_csv(\"/kaggle/input/birdclef-2023/sample_submission.csv\")\nsample_sub[competition_classes] = sample_sub[competition_classes].astype(np.float32)\nsample_sub.head()","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.719306Z","iopub.execute_input":"2023-05-26T13:42:37.719670Z","iopub.status.idle":"2023-05-26T13:42:37.806739Z","shell.execute_reply.started":"2023-05-26T13:42:37.719646Z","shell.execute_reply":"2023-05-26T13:42:37.805129Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"frame_limit_secs = 15 if sample_sub.shape[0] == 3 else None\nfor sample_filename in test_samples:\n    predict_for_sample(sample_filename, sample_sub, frame_limit_secs=15)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:37.809284Z","iopub.execute_input":"2023-05-26T13:42:37.809592Z","iopub.status.idle":"2023-05-26T13:42:59.938097Z","shell.execute_reply.started":"2023-05-26T13:42:37.809570Z","shell.execute_reply":"2023-05-26T13:42:59.937114Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sample_sub.to_csv(\"submission.csv\", index=False)","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:59.939467Z","iopub.execute_input":"2023-05-26T13:42:59.940111Z","iopub.status.idle":"2023-05-26T13:42:59.957706Z","shell.execute_reply.started":"2023-05-26T13:42:59.940076Z","shell.execute_reply":"2023-05-26T13:42:59.956647Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"sample_sub.head()","metadata":{"execution":{"iopub.status.busy":"2023-05-26T13:42:59.958880Z","iopub.execute_input":"2023-05-26T13:42:59.959238Z","iopub.status.idle":"2023-05-26T13:42:59.986532Z","shell.execute_reply.started":"2023-05-26T13:42:59.959207Z","shell.execute_reply":"2023-05-26T13:42:59.985638Z"},"trusted":true},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"","metadata":{},"execution_count":null,"outputs":[]}]}