UNIMPLEMENTED. Track matching model metrics for a single track match threshold and multiple label match confidence thresholds.
confidenceMetrics[]
object ( ConfidenceMetrics
)
Metrics for each label-match confidenceThreshold
from 0.05,0.10,...,0.95,0.96,0.97,0.98,0.99. Precision-recall curve is derived from them.
iouThreshold
number
The intersection-over-union threshold value between bounding boxes across frames used to compute this metric entry.
meanTrackingAveragePrecision
number
The mean average precision over all confidence thresholds.
meanBoundingBoxIou
number
The mean bounding box iou over all confidence thresholds.
meanMismatchRate
number
The mean mismatch rate over all confidence thresholds.
| JSON representation |
|---|
{
"confidenceMetrics"
:
[
{
object (
|
ConfidenceMetrics
Metrics for a single confidence threshold.
confidenceThreshold
number
The confidence threshold value used to compute the metrics.
trackingPrecision
number
Tracking precision.
trackingRecall
number
Tracking recall.
boundingBoxIou
number
Bounding box intersection-over-union precision. Measures how well the bounding boxes overlap between each other (e.g. complete overlap or just barely above iouThreshold).
mismatchRate
number
Mismatch rate, which measures the tracking consistency, i.e. correctness of instance id continuity.
| JSON representation |
|---|
{ "confidenceThreshold" : number , "trackingPrecision" : number , "trackingRecall" : number , "boundingBoxIou" : number , "mismatchRate" : number } |
