seisbench-model-api

To apply pretrained models to seismic waveforms, load the model via the from_pretrained utility and run the annotate function on an obspy stream to obtain waveform annotations for phase picking.

The SeisBench WaveformModel API provides a structured interface to train and apply machine learning models to seismic data, enabling seamless integration with pretrained weights for earthquake detection and classification workflows.

Yes, you can use pretrained SeisBench models for denoising and earthquake detection by loading the weights and applying the annotate function to convert and process your obspy streams.

You load pretrained weights for seismic data processing using the from_pretrained utility provided by the WaveformModel API, which handles the integration of model weights for tasks like phase picking.

The best way to integrate machine learning models into a seismic data pipeline is using the abstract WaveformModel interface, which ensures consistent model loading, annotation, and evaluation across real-world workflows.

The WaveformModel API requires checking for dependencies and components before annotating seismic data, and it relies on the availability of pretrained weights to successfully execute phase picking and denoising tasks.