Traffic Sign Recognition based on Synthesised Training Data

Stergiou, Alexandros; Kalliatakis, Grigorios; Chrysoulas, Christos

doi:10.3390/bdcc2030019

Cited by 9 publications

(2 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Genetic methods, for example, are used to mutate and crossover existing data to create new data to expand categorisation data and data to help with the recognition of sign language [Wang et al 2006]. Other systems have used neural networks to partly synthesise data by adding synthesised elements to images in order to make early training easier [Stergiou et al 2018]. Very few systems generate entirely new data, though there are some examples of generating meta-data as a new data set to work with instead of the original visual data set [Long et al 2018].…”

Section: Related Workmentioning

confidence: 99%

Code and Data Synthesis for Genetic Improvement in Emergent Software Systems

Rainford

Porter

2022

ACM Trans. Evol. Learn. Optim.

View full text Add to dashboard Cite

Emergent software systems are assembled from a collection of small code blocks, where some of those blocks have alternative implementation variants; they optimise at run-time by learning which compositions of alternative blocks best suit each deployment environment encountered. In this paper we study the automated synthesis of new implementation variants for a running system using genetic improvement (GI) . Typical GI approaches, however, rely on large amounts of data for accurate training and large code bases from which to source genetic material. In emergent systems we have neither asset, with sparsely sampled runtime data and small code volumes in each building block. We therefore examine two approaches to more effective GI under these constraints: the synthesis of data from sparse samples to construct statistically representative larger training corpora; and the synthesis of code to counter the relative lack of genetic material in our starting population members. Our results demonstrate that a mixture of synthesised and existing code is a viable optimisation strategy, and that phases of increased synthesis can make GI more robust to deleterious mutations. On synthesised data, we find that we can produce equivalent optimisation compared to GI methods using larger data sets, and that this optimisation can produce both useful specialists and generalists.

show abstract

Section: Related Workmentioning

confidence: 99%

Code and Data Synthesis for Genetic Improvement in Emergent Software Systems

Rainford

Porter

2022

ACM Trans. Evol. Learn. Optim.

View full text Add to dashboard Cite

show abstract

“…3D models have also been used to generate data for training, overlaying 2D renderings of those models on real images [29]. In the task of traffic sign classification, some works used image processing to generate training samples from traffic sign templates [30], [31]. In the detection task, Møgelmose et al [32] tried using synthetic data to train a Viola-Jones traffic sign detector, but the results were not satisfactory.…”

Section: B Synthetic Data Generation For Deep Trainingmentioning

confidence: 99%

Deep Traffic Sign Detection and Recognition Without Target Domain Real Images

Tabelini¹,

Berriel²,

Paixão³

et al. 2020

Preprint

View full text Add to dashboard Cite

Deep learning has been successfully applied to several problems related to autonomous driving, often relying on large databases of real target-domain images for proper training. The acquisition of such real-world data is not always possible in the self-driving context, and sometimes their annotation is not feasible. Moreover, in many tasks, there is an intrinsic data imbalance that most learning-based methods struggle to cope with. Particularly, traffic sign detection is a challenging problem in which these three issues are seen altogether. To address these challenges, we propose a novel database generation method that requires only (i) arbitrary natural images, i.e., requires no real image from the target-domain, and (ii) templates of the traffic signs. The method does not aim at overcoming the training with real data, but to be a compatible alternative when the real data is not available. The effortlessly generated database is shown to be effective for the training of a deep detector on traffic signs from multiple countries. On large data sets, training with a fully synthetic data set almost matches the performance of training with a real one. When compared to training with a smaller data set of real images, training with synthetic images increased the accuracy by 12.25%. The proposed method also improves the performance of the detector when target-domain data are available.

show abstract

Traffic Sign Repositories: Bridging the Gap Between Real and Synthetic Data

Silva

Fernandes

2023

Communications in Computer and Information Science

View full text Add to dashboard Cite

Traffic Sign Recognition based on Synthesised Training Data

Cited by 9 publications

References 23 publications

Code and Data Synthesis for Genetic Improvement in Emergent Software Systems

Code and Data Synthesis for Genetic Improvement in Emergent Software Systems

Deep Traffic Sign Detection and Recognition Without Target Domain Real Images

Traffic Sign Repositories: Bridging the Gap Between Real and Synthetic Data

Contact Info

Product

Resources

About