evaluateDetectionMissRate

(To be removed) Evaluate miss rate metric for object detection

collapse all in page

evaluateDetectionMissRate will be removed in a future release. Use evaluateObjectDetection instead.

Syntax

logAverageMissRate = evaluateDetectionMissRate(detectionResults,groundTruthData)
[logAverageMissRate,fppi,missRate] = evaluateDetectionMissRate(___)
[___] = evaluateDetectionMissRate(___,threshold)

Description

logAverageMissRate = evaluateDetectionMissRate(detectionResults,groundTruthData) returns the log-average miss rate of the detectionResults compared to groundTruthData, which is used to measure the performance of the object detector. For a multiclass detector, the log-average miss rate is a vector of scores for each object class in the order specified by groundTruthData.

[logAverageMissRate,fppi,missRate] = evaluateDetectionMissRate(___) returns data points for plotting the log miss rate–false positives per image (FPPI) curve, using input arguments from the previous syntax.

[___] = evaluateDetectionMissRate(___,threshold) specifies the overlap threshold for assigning a detection to a ground truth box.

Input Arguments

collapse all

Object locations and scores, specified as a two-column table containing the bounding boxes and scores for each detected object. For multiclass detection, a third column contains the predicted label for each detection. The bounding boxes must be stored in an M-by-4 cell array. The scores must be stored in an M-by-1 cell array, and the labels must be stored as a categorical vector.

When detecting objects, you can create the detection results table by using imageDatastore.

        ds = imageDatastore(stopSigns.imageFilename);
        detectionResults = detect(detector,ds);

Data Types: table

Labeled ground truth, specified as a datastore or a table.

Each bounding box must be in the format [x y width height].

  • Datastore — A datastore whose read and readall functions return a cell array or a table with at least two columns of bounding box and labels cell vectors. The bounding boxes must be in a cell array of M-by-4 matrices in the format [x,y,width,height]. The datastore's read and readall functions must return one of the formats:

    • {boxes,labels} — The boxLabelDatastore creates this type of datastore.

    • {images,boxes,labels} — A combined datastore. For example, using combine(imds,blds).

    See boxLabelDatastore.

  • Table — One or more columns. All columns contain bounding boxes. Each column must be a cell vector that contains M-by-4 matrices that represent a single object class, such as stopSign, carRear, or carFront. The columns contain 4-element double arrays of M bounding boxes in the format [x,y,width,height]. The format specifies the upper-left corner location and size of the bounding box in the corresponding image.

Overlap threshold for a detection assigned to a ground truth box, specified as a numeric scalar. The overlap ratio is computed as the intersection over union.

Output Arguments

collapse all

Log-average miss rate metric, returned as either a numeric scalar or vector. For a multiclass detector, the log-average miss rate is returned as a vector of values that correspond to the data points for each class.

False positives per image, returned as either a vector of numeric scalars or as a cell array. For a multiclass detector, the FPPI and log miss rate are cell arrays, where each cell contains the data points for each object class.

Log miss rate, returned as either a vector of numeric scalars or as a cell array. For a multiclass detector, the FPPI and log miss rate are cell arrays, where each cell contains the data points for each object class.

Version History

Introduced in R2017a

expand all

See Also

Apps

Functions

Objects

Topics