-
Notifications
You must be signed in to change notification settings - Fork 0
/
Evaluation.m
48 lines (43 loc) · 1.57 KB
/
Evaluation.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
clear;
load('AllProbabilityMaps.mat');
file = fopen('config.txt');
DataLocation = fgetl(file);
[~, ~, ~, test_masks] = partitionDataset(DataLocation);
ObjectCount = 8;
threshold = 0.6;
% Rows are objects and columns are Positives Correctly Detected, Total
% Positives, Negatives Incorrectly Detected, Total Negatives respectively
% Last two columns are TPR and FPR
ROC = zeros(ObjectCount,6);
for i = 1:size(test_masks, 2)
current_mask = test_masks{1,i};
num_of_objects = length(current_mask);
for j = 1:num_of_objects
className = current_mask(j).class_name;
mask = current_mask(j).mask;
classId = getClassId(className);
probMask = AllProbabilityMaps{i,1}{1,classId};
decisionMask = probMask > threshold;
height = size(mask, 1);
width = size(mask, 2);
for m = 1:height
for n = 1:width
if mask(m,n) == 1 % positive pixel
ROC(classId,2) = ROC(classId,2) + 1;
if decisionMask(m,n) == 1 % true positive
ROC(classId,1) = ROC(classId,1) + 1;
end
elseif mask(m,n) == 0 % negative pixel
ROC(classId,4) = ROC(classId,4) + 1;
if decisionMask(m,n) == 1 % incorrect negative
ROC(classId,3) = ROC(classId,3) + 1;
end
end
end
end
end
end
for i = 1:ObjectCount
ROC(i,5) = ROC(i,1)/ROC(i,2); % true positive rate (TPR)
ROC(i,6) = ROC(i,3)/ROC(i,4); % false positive rate (FPR)
end