nicetoolbox.detectors.feature_detectors.base_feature

Base class for Feature Detectors. Feature detectors run computations in-process using method detector outputs.

Classes

BaseFeature

Abstract base class for feature detectors.
class nicetoolbox.detectors.feature_detectors.base_feature.BaseFeature(io, data, sequence_context, algorithm_instance: str)[source]

Abstract base class for feature detectors.

Feature detectors run in-process, computing derived features from method detector outputs.

Initialize base detector with references.

Subclasses should call super().__init__() and set inference_config.

abstract compute() → Any[source]

Compute the feature from method detector outputs.

Returns:

Computed feature data (passed to visualization and post_compute)

get_input_file(component: str, algorithm: str) → Path[source]

Get the input file path for a specific upstream detector.

Parameters:

  • component – Component name (e.g., ‘body_joints’)

  • algorithm – Algorithm name (e.g., ‘hrnetw48’)

Returns:

Path to the .npz result file

run() → Any[source]

Execute feature detector: compute() + post_compute().

Returns computed data for visualization.