reidentificationMetrics

Re-identification (ReID) quality metrics

Since R2024a

Description

An reidentificationMetrics object stores ReID quality metrics, such as the cumulative matching characteristic (CMC) curve and mean average precision (mAP), for a set of images.

Creation

Create an reidentificationMetrics object by using the evaluateReidentificationNetwork function.

Properties

DataSetMetrics — Metrics aggregated over the data set

Read-only: table

This property is read-only.

Metrics aggregated over the data set, stored as a table with one row. The DatasetMetrics property has columns corresponding to these evaluation metrics. These metrics are not in output table order.

• AP — Average precision, stored as a numeric scalar. The average precision metric represents the accuracy of image search results for a query image.

• mAP — Mean average precision (mAP), stored as a numeric scalar. The mAP is an average of the average precision AP in the same table across all corresponding queries.

• CMC — Cumulative matching characteristic, stored as a numeric scalar.

ClassMetrics — Metrics for each class

Read-only: table

This property is read-only.

Metrics for each class, stored as a table with C rows, where C is the number of object classes specified to the labels argument of the evaluateReidentificationNetwork function. The ClassMetrics has columns corresponding to these evaluation metrics. These metrics are not in output table order.

• AP — Average precision, stored as a numeric scalar. The average precision metric represents the accuracy of image search results for a query image.

• mAP — Mean average precision (mAP), stored as a numeric scalar. The mAP is an average of the average precision AP in the same table across all corresponding queries.

• CMC — Cumulative matching characteristic, stored as a numeric scalar.

ClassNames — Object class names

Read-only: cell array | categorical vector | cell array of character vectors

This property is read-only.

Names of the object classes that the ReID network is trained to re-identify, stored as a cell array, categorical vector, or cell array of character vectors.

Object Functions

plot	Plot re-identification (ReID) metrics

Version History

Introduced in R2024a