2018 IEEE Intelligent Vehicles Symposium (IV) 2018
DOI: 10.1109/ivs.2018.8500497
|View full text |Cite
|
Sign up to set email alerts
|

A New Metric for Evaluating Semantic Segmentation: Leveraging Global and Contour Accuracy

Abstract: Semantic segmentation of images is an important problem for mobile robotics and autonomous driving because it offers basic information which can be used for complex reasoning and safe navigation. Different solutions have been proposed for this problem along the last two decades, and a relevant increment on accuracy has been achieved recently with the application of deep neural networks for image segmentation. One of the main issues when comparing different neural networks architectures is how to select an appr… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
45
0

Year Published

2019
2019
2024
2024

Publication Types

Select...
6
4

Relationship

0
10

Authors

Journals

citations
Cited by 78 publications
(45 citation statements)
references
References 21 publications
0
45
0
Order By: Relevance
“…We evaluated the performance of our model using four metrics: accuracy score, precision, recall [31], and f 1 -score [32]. The accuracy score is the ratio of the correctly classified elements over all available elements.…”
Section: Evaluation Metricsmentioning
confidence: 99%
“…We evaluated the performance of our model using four metrics: accuracy score, precision, recall [31], and f 1 -score [32]. The accuracy score is the ratio of the correctly classified elements over all available elements.…”
Section: Evaluation Metricsmentioning
confidence: 99%
“…The possibility of having a lesion absence in the image can be presented, resulting in the consideration of patients as normal and healthy, therefore; no segmentation is required. Besides, we assessed statistically the segmentation quality of the proposed against GraphCut, Watershed, and MSA, by picking: Accuracy [37] , Sensitivity [38] , F-Measure [38] , Precision [39] , MCC (Mathew Correlation Coefficient) [38] , Dice [40] , Jaccard [40] , and Specificity [38] . By definition, higher values on these indexes imply a better quality of segmentation.…”
Section: Resultsmentioning
confidence: 99%
“…Standard accuracy metrics for the tasks in the domain of computer vision are accuracy, precision, recall, and Fmeasure [75,76]. For the object detection task, the mean average precision (mAP) measure is also used to evaluate the performance of the models [77].…”
Section: B Evaluation Measurementioning
confidence: 99%