API Reference

Pipeline module for the nod detection system. This module contains the base pipeline and related components.

class nod_detector.pipeline.BasePipeline

Bases: ABC, Generic[T]

Abstract base class for processing pipelines.

This class defines the interface that all pipeline implementations should follow.

Type Variables:

T: The return type of the _process method, must be a dictionary.

__init__() → None

Initialize the pipeline with default parameters.

property is_running: bool

Return whether the pipeline is currently running.

Returns:

True if the pipeline is running, False otherwise.

Return type:

bool

process(input_data: Any, **kwargs: Any) → T

Process the input data through the pipeline.

Parameters:

• input_data – The input data to process.

• **kwargs – Additional parameters specific to the pipeline implementation.

Returns:

The processing results as a dictionary.

Raises:

RuntimeError – If the pipeline is already running.

reset() → None

Reset the pipeline state.

class nod_detector.pipeline.VideoProcessingPipeline(config: Dict[str, Any] | None = None)

Bases: BasePipeline[ProcessingResults]

Pipeline for processing video frames for nod detection.

This pipeline handles: - Video frame extraction - Frame preprocessing - Person detection (to be implemented) - Head pose estimation (to be implemented) - Nod detection (to be implemented) - Result visualization using rerun.io

__init__(config: Dict[str, Any] | None = None) → None

Initialize the video processing pipeline.

Parameters:

config – Configuration dictionary for the pipeline.

reset() → None

Reset the pipeline state.