Evaluation of deep neural networks for traffic sign detection systems

Arcos-García, Álvaro; Álvarez-García, Juan Antonio; Morillo, Luis Miguel Soria

doi:10.1016/j.neucom.2018.08.009

Cited by 194 publications

(97 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The traffic sign detection and recognition problem has been investigated by the research community for a while. Researchers have been proposing all types of solutions such as the ones using hand-crafted features in model-based solutions [9], leveraging simple features in learning-based approaches [10], and, the more recent and state-of-the-art, using deep learning based methods [6], [11] that is the focus of this work. A detailed and complete review of these methods can be found in [12].…”

Section: Introductionmentioning

confidence: 99%

Effortless Deep Training for Traffic Sign Detection Using Templates and Arbitrary Natural Images

Tabelini

Paixão

Berriel

et al. 2019

2019 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

Deep learning has been successfully applied to several problems related to autonomous driving. Often, these solutions rely on large networks that require databases of real image samples of the problem (i.e., real world) for proper training. The acquisition of such real-world data sets is not always possible in the autonomous driving context, and sometimes their annotation is not feasible (e.g., takes too long or is too expensive). Moreover, in many tasks, there is an intrinsic data imbalance that most learning-based methods struggle to cope with. It turns out that traffic sign detection is a problem in which these three issues are seen altogether. In this work, we propose a novel database generation method that requires only (i) arbitrary natural images, i.e., requires no real image from the domain of interest, and (ii) templates of the traffic signs, i.e., templates synthetically created to illustrate the appearance of the category of a traffic sign. The effortlessly generated training database is shown to be effective for the training of a deep detector (such as Faster R-CNN) on German traffic signs, achieving 95.66% of mAP on average. In addition, the proposed method is able to detect traffic signs with an average precision, recall and F1-score of about 94%, 91% and 93%, respectively. The experiments surprisingly show that detectors can be trained with simple data generation methods and without problem domain data for the background, which is in the opposite direction of the common sense for deep learning.

show abstract

Section: Introductionmentioning

confidence: 99%

Effortless Deep Training for Traffic Sign Detection Using Templates and Arbitrary Natural Images

Tabelini

Paixão

Berriel

et al. 2019

2019 International Joint Conference on Neural Networks (IJCNN)

View full text Add to dashboard Cite

show abstract

“…8. Faster R-CNN Inception Resnet V2 model architecture [53] for iris certification. The CRR of the three iris libraries compared to the existing algorithms are shown in Table 9.…”

Section: Existing Methods Comparison Experimentsmentioning

confidence: 99%

Unsteady State Lightweight Iris Certification Based on Multi-Algorithm Parallel Integration

Liu

Zhu

et al. 2019

Algorithms

View full text Add to dashboard Cite

Aimed at the one-to-one certification problem of unsteady state iris at different shooting times, a multi-algorithm parallel integration general model structure is proposed in this paper. The iris in the lightweight constrained state affected by defocusing, deflection, and illumination is taken as the research object, the existing algorithms are combined into the model structure effectively, and a one-to-one certification algorithm for lightweight constrained state unsteady iris was designed based on multi-algorithm integration and maximum trusted decision. In this algorithm, a sufficient number of iris internal feature points from the unstable state texture were extracted as effective iris information through the image processing layer composed of various filtering processing algorithms, thereby eliminating defocused interference. In the feature recognition layer, iris deflection interference was excluded by the improved methods of Gabor and Hamming and Haar and BP for the stable features extracted by the image processing layer, and two certification results were obtained by means of parallel recognition. The correct number of certifications for an algorithm under a certain lighting condition were counted. The method with the most correct number was set as the maximum trusted method under this lighting condition, and the results of the maximum trusted method were taken as the final decision, thereby eliminating the effect of illumination. Experiments using the JLU and CASIA iris libraries under the prerequisites in this paper show that the correct recognition rate of the algorithm can reach a high level of 98% or more, indicating that the algorithm can effectively improve the accuracy of the one-to-one certification of lightweight constrained state unsteady iris. Compared with the latest architecture algorithms, such as CNN and deep learning, the proposed algorithm is more suitable for the prerequisites presented in this paper, which has good environmental inclusiveness and can better improve existing traditional algorithms’ effectiveness through the design of a parallel integration model structure.

show abstract

“…By these methods a feature vector is extracted from image with lower computational cost than from 3-D point cloud. Then, the class label of the feature vector is obtained using a classifier such as Support Vector Machine (SVM) or with Deep Learning-based (DL) methods [42][43][44]. Among the latter, Convolutional Neural Networks (CNN) have been widely adopted, given their high performance in both TSD and TSR in images [45][46][47][48] and in point clouds [49].…”

Section: Related Workmentioning

confidence: 99%

Novel Approach to Automatic Traffic Sign Inventory Based on Mobile Mapping System Data and Deep Learning

et al. 2020

View full text Add to dashboard Cite

Traffic signs are a key element in driver safety. Governments invest a great amount of resources in maintaining the traffic signs in good condition, for which a correct inventory is necessary. This work presents a novel method for mapping traffic signs based on data acquired with MMS (Mobile Mapping System): images and point clouds. On the one hand, images are faster to process and artificial intelligence techniques, specifically Convolutional Neural Networks, are more optimized than in point clouds. On the other hand, point clouds allow a more exact positioning than the exclusive use of images. The false positive rate per image is only 0.004. First, traffic signs are detected in the images obtained by the 360° camera of the MMS through RetinaNet and they are classified by their corresponding InceptionV3 network. The signs are then positioned in the georeferenced point cloud by means of a projection according to the pinhole model from the images. Finally, duplicate geolocalized signs detected in multiple images are filtered. The method has been tested in two real case studies with 214 images, where 89.7% of the signals have been correctly detected, of which 92.5% have been correctly classified and 97.5% have been located with an error of less than 0.5 m. This sequence, which combines images to detection–classification, and point clouds to geo-referencing, in this order, optimizes processing time and allows this method to be included in a company’s production process. The method is conducted automatically and takes advantage of the strengths of each data type.

show abstract

Evaluation of deep neural networks for traffic sign detection systems

Cited by 194 publications

References 33 publications

Effortless Deep Training for Traffic Sign Detection Using Templates and Arbitrary Natural Images

Effortless Deep Training for Traffic Sign Detection Using Templates and Arbitrary Natural Images

Unsteady State Lightweight Iris Certification Based on Multi-Algorithm Parallel Integration

Novel Approach to Automatic Traffic Sign Inventory Based on Mobile Mapping System Data and Deep Learning

Contact Info

Product

Resources

About