{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.12","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"},"kaggle":{"accelerator":"none","dataSources":[{"sourceId":44224,"databundleVersionId":5188730,"sourceType":"competition"},{"sourceId":3836,"sourceType":"modelInstanceVersion","modelInstanceId":2739,"modelId":319}],"dockerImageVersionId":30407,"isInternetEnabled":false,"language":"python","sourceType":"notebook","isGpuEnabled":false}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"markdown","source":"In this notebook, we'll use a pre-trained machine learning model to generate a submission to the [BirdClef2023 competition](https://www.kaggle.com/c/birdclef-2023).  The goal of the competition is to identify Eastern African bird species by sound.","metadata":{}},{"cell_type":"markdown","source":"## Step 1: Imports","metadata":{}},{"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":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2025-01-08T16:28:27.324345Z","iopub.execute_input":"2025-01-08T16:28:27.324902Z","iopub.status.idle":"2025-01-08T16:28:39.902271Z","shell.execute_reply.started":"2025-01-08T16:28:27.324836Z","shell.execute_reply":"2025-01-08T16:28:39.900506Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"## Step 2: Explore the training data\n\nWe'll start by loading a couple of training examples and using the IPython.display.Audio module to play them!","metadata":{}},{"cell_type":"code","source":"# Load a sample audio files from two different species\naudio_abe, sr_abe = librosa.load(\"/kaggle/input/birdclef-2023/train_audio/abethr1/XC128013.ogg\")\naudio_abh, sr_abh = librosa.load(\"/kaggle/input/birdclef-2023/train_audio/abhori1/XC127317.ogg\")","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:28:39.905701Z","iopub.execute_input":"2025-01-08T16:28:39.906528Z","iopub.status.idle":"2025-01-08T16:28:53.365125Z","shell.execute_reply.started":"2025-01-08T16:28:39.906435Z","shell.execute_reply":"2025-01-08T16:28:53.363448Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"# Play the audio\nAudio(data=audio_abe, rate=sr_abe)","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:28:53.366984Z","iopub.execute_input":"2025-01-08T16:28:53.368065Z","iopub.status.idle":"2025-01-08T16:28:53.457249Z","shell.execute_reply.started":"2025-01-08T16:28:53.367999Z","shell.execute_reply":"2025-01-08T16:28:53.454776Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"# Play the audio\nAudio(data=audio_abh, rate=sr_abh)","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:28:53.459211Z","iopub.execute_input":"2025-01-08T16:28:53.459712Z","iopub.status.idle":"2025-01-08T16:28:53.540780Z","shell.execute_reply.started":"2025-01-08T16:28:53.459643Z","shell.execute_reply":"2025-01-08T16:28:53.539089Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"## Step 3: Match the model's output with the bird species in the competition\n\nThe competition includes 264 classes of birds, 261 of which exist in this model. We'll set up a way to map the model's output logits to our competition.","metadata":{}},{"cell_type":"code","source":"model = hub.load('https://kaggle.com/models/google/bird-vocalization-classifier/frameworks/tensorFlow2/variations/bird-vocalization-classifier/versions/1')\nlabels_path = hub.resolve('https://kaggle.com/models/google/bird-vocalization-classifier/frameworks/tensorFlow2/variations/bird-vocalization-classifier/versions/1') + \"/assets/label.csv\"","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:28:53.542574Z","iopub.execute_input":"2025-01-08T16:28:53.543996Z","iopub.status.idle":"2025-01-08T16:29:02.027375Z","shell.execute_reply.started":"2025-01-08T16:28:53.543946Z","shell.execute_reply":"2025-01-08T16:29:02.025548Z"},"trusted":true},"outputs":[],"execution_count":null},{"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## note that the bird classifier classifies a much larger set of birds than the\n## competition, so we need to load the model's set of class names or else our \n## indices will be off.\nclasses = class_names_from_csv(labels_path)","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:02.029958Z","iopub.execute_input":"2025-01-08T16:29:02.030386Z","iopub.status.idle":"2025-01-08T16:29:02.048079Z","shell.execute_reply.started":"2025-01-08T16:29:02.030345Z","shell.execute_reply":"2025-01-08T16:29:02.045950Z"},"trusted":true},"outputs":[],"execution_count":null},{"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## this is the count of classes not supported by our pretrained model\n## you could choose to simply not predict these, set a default as above,\n## or create your own model using the pretrained model as a base.\nforced_defaults","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:02.052735Z","iopub.execute_input":"2025-01-08T16:29:02.053160Z","iopub.status.idle":"2025-01-08T16:29:02.267709Z","shell.execute_reply.started":"2025-01-08T16:29:02.053121Z","shell.execute_reply":"2025-01-08T16:29:02.266255Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"## Step 4: Preprocess the data\n\nThe following functions are one way to load the audio provided and break it up into the five-second samples with a sample rate of 32,000 required by the competition.","metadata":{}},{"cell_type":"code","source":"def 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":"2025-01-08T16:29:02.269514Z","iopub.execute_input":"2025-01-08T16:29:02.270005Z","iopub.status.idle":"2025-01-08T16:29:02.279994Z","shell.execute_reply.started":"2025-01-08T16:29:02.269951Z","shell.execute_reply":"2025-01-08T16:29:02.278488Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"Below we load one training sample - use the Audio function to listen to the samples inside the notebook!","metadata":{}},{"cell_type":"code","source":"audio, sample_rate = librosa.load(\"/kaggle/input/birdclef-2023/train_audio/afghor1/XC156639.ogg\")\nsample_rate, wav_data = ensure_sample_rate(audio, sample_rate)\nAudio(wav_data, rate=sample_rate)","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:02.282282Z","iopub.execute_input":"2025-01-08T16:29:02.283869Z","iopub.status.idle":"2025-01-08T16:29:03.114255Z","shell.execute_reply.started":"2025-01-08T16:29:02.283795Z","shell.execute_reply":"2025-01-08T16:29:03.111680Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"## Step 5: Make predictions\n\nEach test sample is cut into 5-second chunks. We use the pretrained model to return probabilities for all 10k birds included in the model, then pull out the classes used in this competition to create a final submission row. Note that we are NOT doing anything special to handle the 3 missing classes; those will need fine-tuning / transfer learning, which will be handled in a separate notebook.","metadata":{}},{"cell_type":"code","source":"fixed_tm = frame_audio(wav_data)\nlogits, embeddings = model.infer_tf(fixed_tm[:1])\nprobabilities = tf.nn.softmax(logits)\nargmax = np.argmax(probabilities)\nprint(f\"The audio is from the class {classes[argmax]} (element:{argmax} in the label.csv file), with probability of {probabilities[0][argmax]}\")","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:03.116170Z","iopub.execute_input":"2025-01-08T16:29:03.116557Z","iopub.status.idle":"2025-01-08T16:29:15.318689Z","shell.execute_reply.started":"2025-01-08T16:29:03.116517Z","shell.execute_reply":"2025-01-08T16:29:15.317146Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"def 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":"2025-01-08T16:29:15.320359Z","iopub.execute_input":"2025-01-08T16:29:15.320868Z","iopub.status.idle":"2025-01-08T16:29:15.332071Z","shell.execute_reply.started":"2025-01-08T16:29:15.320816Z","shell.execute_reply":"2025-01-08T16:29:15.330648Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"markdown","source":"## Step 6: Generate a submission\n\nNow we process all of the test samples as discussed above, creating output rows, and saving them in the provided `sample_submission.csv`. Finally, we save these rows to our final output file: `submission.csv`. This is the file that gets submitted and scored when you submit the notebook.","metadata":{}},{"cell_type":"code","source":"test_samples = list(glob.glob(\"/kaggle/input/birdclef-2023/test_soundscapes/*.ogg\"))\ntest_samples","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:15.333963Z","iopub.execute_input":"2025-01-08T16:29:15.334403Z","iopub.status.idle":"2025-01-08T16:29:15.375613Z","shell.execute_reply.started":"2025-01-08T16:29:15.334363Z","shell.execute_reply":"2025-01-08T16:29:15.374201Z"},"trusted":true},"outputs":[],"execution_count":null},{"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":"2025-01-08T16:29:15.377101Z","iopub.execute_input":"2025-01-08T16:29:15.377430Z","iopub.status.idle":"2025-01-08T16:29:15.491116Z","shell.execute_reply.started":"2025-01-08T16:29:15.377399Z","shell.execute_reply":"2025-01-08T16:29:15.489675Z"},"trusted":true},"outputs":[],"execution_count":null},{"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":"2025-01-08T16:29:15.492680Z","iopub.execute_input":"2025-01-08T16:29:15.493095Z","iopub.status.idle":"2025-01-08T16:29:28.766809Z","shell.execute_reply.started":"2025-01-08T16:29:15.493059Z","shell.execute_reply":"2025-01-08T16:29:28.765029Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"sample_sub","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:28.768654Z","iopub.execute_input":"2025-01-08T16:29:28.769161Z","iopub.status.idle":"2025-01-08T16:29:28.801506Z","shell.execute_reply.started":"2025-01-08T16:29:28.769106Z","shell.execute_reply":"2025-01-08T16:29:28.799910Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"sample_sub.to_csv(\"submission.csv\", index=False)","metadata":{"execution":{"iopub.status.busy":"2025-01-08T16:29:28.803281Z","iopub.execute_input":"2025-01-08T16:29:28.803867Z","iopub.status.idle":"2025-01-08T16:29:28.827015Z","shell.execute_reply.started":"2025-01-08T16:29:28.803829Z","shell.execute_reply":"2025-01-08T16:29:28.825513Z"},"trusted":true},"outputs":[],"execution_count":null},{"cell_type":"code","source":"","metadata":{"trusted":true},"outputs":[],"execution_count":null}]}