nicetoolbox.evaluation.metrics.joints.missing_points

Classes

MissingPointsMetric

Detect missing joints/landmarks per frame as a binary flag.
class nicetoolbox.evaluation.metrics.joints.missing_points.MissingPointsMetric(metrics_config: BaseMetricConfig, config_handler: ConfigHandler)[source]

Detect missing joints/landmarks per frame as a binary flag.

A point is considered missing when its spatial coordinates are all NaN. Optionally, points whose confidence score falls below min_confidence are also treated as missing (requires a confidence in axis4).

Output shape: (subjects, cameras, frames, labels) — float32, 1.0 = missing, 0.0 = present. Summary and score are expressed as detection rate (1 - missing rate), so higher is better.

Parameters:

  • metrics_config – Metric-specific configuration.

  • config_handler – Shared config handler with project and evaluation configs.

compute() → MetricResult[source]

Execute the metric end-to-end: load data, compute, return results.

Returns:

MetricResult containing summary tables, per-frame arrays, and plots.