This figure presents the confusion matrices computed on the whole test set for three probability thresholds value: (a) T = 0.2, (b) T = 0.4 and (c) T = 0.6, as defined in the Results section. These correspond to a more detailed and traditional version of those presented in the paper (Fig 3), with corresponding surfaces (in ha) and accompanied with some metrics values (accuracy, precision, recall and F1-score). (TIF)</p
Results are averaged across observers and execution of transformation and viewpoint. Actual transfor...
<p>Confusion matrix between the observed and predicted values of four hardness classes based on the ...
<p>The confusion matrices of a) Tabassian’s method, and b) proposed method for the artificial datase...
<p>(A), Confusion matrix of the model performance on the dataset of 15 scenes. The average accuracy ...
The colours of the heat map correspond to the percentage of classification in each category. The acc...
Actual materials on the y axis are plotted against perceived materials on the x axis. R2 values indi...
<p>The numbers on the diagonal indicate the predictive accuracy (in %) of the stBF-classifier for ea...
(a) Recovery rates across families. Rows represent true models/families (used to simulate data), and...
<p>ACC: accuracy; ERR: error rate; SN: sensitivity; TPR: true positive rate; REC: recall; SP: specif...
There are strong incentives to build classification systems that show outstanding performance on var...
Comparison of confusion matrices between Findaureus and Otsu’s thresholding method findings, as pres...
<p>Estimates across variations of each situation are derived from tow-tailed exact binomial tests (π...
<p>The confusion matrix for discriminating A three alternatives: and B nine alternatives: is shown...
These confusion matrices show the result of applying the empirically-optimal threshold values of 61%...
<p>The data presented in each column is the probability of the participants' judgments. The recognit...
Results are averaged across observers and execution of transformation and viewpoint. Actual transfor...
<p>Confusion matrix between the observed and predicted values of four hardness classes based on the ...
<p>The confusion matrices of a) Tabassian’s method, and b) proposed method for the artificial datase...
<p>(A), Confusion matrix of the model performance on the dataset of 15 scenes. The average accuracy ...
The colours of the heat map correspond to the percentage of classification in each category. The acc...
Actual materials on the y axis are plotted against perceived materials on the x axis. R2 values indi...
<p>The numbers on the diagonal indicate the predictive accuracy (in %) of the stBF-classifier for ea...
(a) Recovery rates across families. Rows represent true models/families (used to simulate data), and...
<p>ACC: accuracy; ERR: error rate; SN: sensitivity; TPR: true positive rate; REC: recall; SP: specif...
There are strong incentives to build classification systems that show outstanding performance on var...
Comparison of confusion matrices between Findaureus and Otsu’s thresholding method findings, as pres...
<p>Estimates across variations of each situation are derived from tow-tailed exact binomial tests (π...
<p>The confusion matrix for discriminating A three alternatives: and B nine alternatives: is shown...
These confusion matrices show the result of applying the empirically-optimal threshold values of 61%...
<p>The data presented in each column is the probability of the participants' judgments. The recognit...
Results are averaged across observers and execution of transformation and viewpoint. Actual transfor...
<p>Confusion matrix between the observed and predicted values of four hardness classes based on the ...
<p>The confusion matrices of a) Tabassian’s method, and b) proposed method for the artificial datase...