BirdSet: A multi-task benchmark and data pipeline for deep learning based avian bioacoustics
Project description
BirdSet Benchmark
Results
Title | Notes | PER | NES | UHH | HSN | NBP | POW | SSW | SNE | Overall | Code |
---|---|---|---|---|---|---|---|---|---|---|---|
BirdSet: A Multi-Task Benchmark For Classification In Avian Bioacoustics | |||||||||||
BIRB: A Generalization Benchmark for Information Retrieval in Bioacoustics |
Get Started
Devcontainer
You can use the devcontainer configured as as git submodule:
git submodule update --init --recursive
Install dependencies
conda create -n birdset python=3.10
pip install -e .
Or poetry.
poetry install
poetry shell
Minimal Working Example
Log in to Huggingface
Our datasets are shared via HuggingFace Datasets in our HuggingFace BirdSet repository. Huggingface is a central hub for sharing and utilizing datasets and models, particularly beneficial for machine learning and data science projects. For accessing private datasets hosted on HuggingFace, you need to be authenticated. Here's how you can log in to HuggingFace:
-
Install HuggingFace CLI: If you haven't already, you need to install the HuggingFace CLI (Command Line Interface). This tool enables you to interact with HuggingFace services directly from your terminal. You can install it using pip:
pip install huggingface_hub
-
Login via CLI: Once the HuggingFace CLI is installed, you can log in to your HuggingFace account directly from your terminal. This step is essential for accessing private datasets or contributing to the HuggingFace community. Use the following command:
huggingface-cli login
After executing this command, you'll be prompted to enter your HuggingFace credentials (User Access Token). Once authenticated, your credentials will be saved locally, allowing seamless access to HuggingFace resources.
Prepare Data
from birdset.datamodule.base_datamodule import DatasetConfig
from birdset.datamodule.birdset_datamodule import BirdSetDataModule
# initiate the data module
dm = BirdSetDataModule(
dataset= DatasetConfig(
data_dir='../../data_birdset/HSN',
dataset_name='HSN',
hf_path='DBD-research-group/BirdSet',
hf_name='HSN',
n_classes=21,
n_workers=3,
val_split=0.2,
task="multilabel",
classlimit=500,
eventlimit=5,
sampling_rate=32000,
),
)
# prepare the data (download dataset, ...)
dm.prepare_data()
# setup the dataloaders
dm.setup(stage="fit")
# get the dataloaders
train_loader = dm.train_dataloader()
Prepare Model and Start Training
from lightning import Trainer
min_epochs = 1
max_epochs = 5
trainer = Trainer(min_epochs=min_epochs, max_epochs=max_epochs, accelerator="gpu", devices=1)
from birdset.modules.base_module import BaseModule
model = BaseModule(
len_trainset=dm.len_trainset,
task=dm.task,
batch_size=dm.train_batch_size,
num_epochs=max_epochs)
trainer.fit(model, dm)
Logging
Logs will be written to Weights&Biases by default.
Background noise
To enhance model performance we mix in additional background noise from downloaded from the DCASE18. To download the files and convert them to the correct format, run the notebook 'download_background_noise.ipynb' in the 'notebooks' folder.
Run experiments
Our experiments are defined in the configs/experiment
folder. To run an experiment, use the following command:
python birdset/main.py experiment=EXPERIMENT_NAME
Project structure
This repository is inspired by the Yet Another Lightning Hydra Template.
├── configs <- Hydra configuration files
│ ├── callbacks <- Callbacks configs
│ ├── datamodule <- Datamodule configs
│ ├── debug <- Debugging configs
│ ├── experiment <- Experiment configs
│ ├── extras <- Extra utilities configs
│ ├── hydra <- Hydra settings configs
│ ├── logger <- Logger configs
│ ├── module <- Module configs
│ ├── paths <- Project paths configs
│ ├── trainer <- Trainer configs
│ ├── transformations <- Transformations / augmentation configs
│ |
│ ├── main.yaml <- Main config
│
├── data_birdset <- Project data
├── dataset <- Code to build the BirdSet dataset
├── notebooks <- Jupyter notebooks.
│
├── birdset <- Source code
│ ├── augmentations <- Augmentations
│ ├── callbacks <- Additional callbacks
│ ├── datamodules <- Lightning datamodules
│ ├── modules <- Lightning modules
│ ├── utils <- Utility scripts
│ │
│ ├── main.py <- Run experiments
│
├── .gitignore <- List of files ignored by git
├── pyproject.toml <- Poetry project file
├── requirements.txt <- File for installing python dependencies
├── requirements-dev.txt <- File for installing python dev dependencies
├── setup.py <- File for installing project as a package
└── README.md
Data pipeline
Our datasets are shared via HuggingFace Datasets in our BirdSet repository. First log in to HuggingFace with:
huggingface-cli login
For a detailed guide to using the BirdSet data pipeline and its many configuration options, see our comprehensive BirdSet Data Pipeline Tutorial.
Datamodule
The datamodules are defined in birdset/datamodule
and configurations are stored under configs/datamodule
.
base_datamodule
is the main class that can be inherited for specific datasets. It is responsible for preparing the data in the function prepare_data
and loading the data in the function setup
. prepare_data
downloads the dataset, applies preprocessing, creates validation splits and saves the data to disk. setup
initiates the dataloaders and configures data transformations.
The following steps are performed in prepare_data
:
- Data is downloaded from HuggingFace Datasets
_load_data
- Data gets preprocessed with
_preprocess_data
- Data is split into train validation and test sets with
_create_splits
- Length of the dataset gets saved to access later
- Data is saved to disk with
_save_dataset_to_disk
The following steps are performed in setup
:
- Data is loaded from disk with
_get_dataset
in which the transforms are applied
Transformations
Data transformations are referred to data transformations that are applied to the data during training. They include e.g. augmentations. The transformations are added to the huggingface dataset with set_transform
.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.