Question

如果可能的话，我想如何对创建的R-CNN对象检测器进行评估和验证阈值，我试图在附加的脚本中使它达到阈值，但它不起作用，我想为得分打下阈值，如下所示0.58分数和bbox不应出现在结果中
我尝试了以下代码，但出现错误

在此输入代码：-


load('gTruth.mat')
smokedetection = selectLabels(gTruth,'alarm');
if ~isfolder(fullfile('EvaluationData'))

    mkdir EvaluationData
    addpath('EvaluationData');

    evaluationData = objectDetectorTrainingData(gTruth,...
    'SamplingFactor',1,'WriteLocation','EvaluationData');
end
imds = imageDatastore(fullfile('EvaluationData'));
numImages = height(evaluationData);
result(numImages,:) = struct('Boxes',[],'Scores',[]);
for i = 1:numImages

    % Read Image
    I = readimage(imds,i); 

    % Detect the object of interest
    [bboxes, scores] = detect(detector,I,'MiniBatchSize', 32);

    % Store result 
    result(i).Boxes = bboxes;
    T = 0.58; % Define threshold here
idx = scores >= T;
    result(i).Scores = scores(idx);

end
% Convert structure to table
results = struct2table(result);
overlap = 0.1;
% Evaluate Metrics
[ap,recall,precision] = evaluateDetectionPrecision(results...
    ,evaluationData(:,2),overlap);
[am,fppi,missRate] = evaluateDetectionMissRate(results,evaluationData(:,2),overlap);

％，这里是我得到的错误


Error using vision.internal.detector.evaluationInputValidation>checkDetectionResultsTable (line 66)
Invalid score value in row 1 of the detection results table: Expected input to be an array with number of
elements equal to 8.
Error in vision.internal.detector.evaluationInputValidation (line 6)
    checkDetectionResultsTable(detectionResults, groundTruth, mfilename);
Error in evaluateDetectionPrecision (line 94)
vision.internal.detector.evaluationInputValidation(detectionResults, ...
Error in Evaluationthedetector (line 33)
[ap,recall,precision] = evaluateDetectionPrecision(results...

置信度阈值

0 个答案: