TY  - CONF
Y1  - 2020///
SN  - 9781728154473
PB  - Institute of Electrical and Electronics Engineers Inc.
UR  - https://www.scopus.com/inward/record.uri?eid=2-s2.0-85097532534&doi=10.1109%2fICCI51257.2020.9247666&partnerID=40&md5=2758293eafa9dbd7d5a0e244112984ec
A1  - Fujita, H.
A1  - Itagaki, M.
A1  - Ichikawa, K.
A1  - Hooi, Y.K.
A1  - Kawahara, K.
A1  - Sarlan, A.
EP  - 22
AV  - none
N1  - cited By 1; Conference of 2020 International Conference on Computational Intelligence, ICCI 2020 ; Conference Date: 8 October 2020 Through 9 October 2020; Conference Code:164916
N2  - This study evaluates road surface object detection tasks using four Mask R-CNN models available on the Tensor-Flow Object Detection API. The models were pre-trained using COCO datasets and fine-tuned by 15,1SS segmented road surface annotation tags. Validation data set was used to obtain Average Precisions and Average Recalls. Result indicates a substantial false negatives or "left judgement"counts for linear cracks, joints, fillings, potholes, stains, shadows and patching with grid cracks classes. There were significant number of incorrectly predicted label instances. To improve the result, an alternative metric calculation method was tested. However, the results showed strong mutual interferences caused by misinterpretation of the scratches with other object classes. © 2020 IEEE.
KW  - Convolutional neural networks; Intelligent computing; Object recognition; Roads and streets
KW  -  CNN models; False negatives; Four Mask; Metric calculation methods; Mutual interference; Object class; Road surfaces; Validation data
KW  -  Object detection
TI  - Fine-tuned Surface Object Detection Applying Pre-trained Mask R-CNN Models
SP  - 17
ID  - scholars12640
ER  -