48,736 Hits in 3.4 sec

Road images augmentation with synthetic traffic signs using neural networks [article]

Anton Konushin, Boris Faizov, Vlad Shakhuro
2021 arXiv   pre-print
We propose three methods for making synthetic signs consistent with a scene in appearance. These methods are based on modern generative adversarial network (GAN) architectures.  ...  We aim to solve that problem by using synthetic training data. Such training data is obtained by embedding synthetic images of signs in the real photos.  ...  Conclusion In this paper, we proposed a neural network-based method for embedding new synthetic traffic signs in road images.  ... 
arXiv:2101.04927v1 fatcat:v5qvswlskrfldeyhxnqzdjvpem

Creating Realistic Power Distribution Networks using Interdependent Road Infrastructure [article]

Rounak Meyur, Madhav Marathe, Anil Vullikanti, Henning Mortveit, Virgilio Centeno, Arun Phadke
2020 arXiv   pre-print
The synthetic network connects high voltage substations to individual residential consumers through primary and secondary distribution networks.  ...  This work proposes a methodology to generate realistic synthetic power distribution networks for a given geographical region.  ...  While the synthetic network almost retraces the road network, the actual network is adjacent to it.  ... 
arXiv:2001.09130v3 fatcat:4o3667vcgravjotoglq4qn4hgq

Road images augmentation with synthetic traffic signs using neural networks

A.S. Konushin, B.V. Faizov, V.I. Shakhuro
2021 Computer Optics  
We propose three methods for making synthetic signs consistent with a scene in appearance. These methods are based on modern generative adversarial network (GAN) architectures.  ...  We aim to solve that problem by using synthetic training data. Such training data is obtained by embedding synthetic images of signs in the real photos.  ...  We also made a comparison of different synthetic data quality by training target neural networks.  ... 
doi:10.18287/2412-6179-co-859 fatcat:grncol6wzvehlb5rcvgxfakery

Semantic Segmentation with Transfer Learning for Off-Road Autonomous Driving

Suvash Sharma, John E. Ball, Bo Tang, Daniel W. Carruth, Matthew Doude, Muhammad Aminul Islam
2019 Sensors  
Second, we used synthetic datasets as the intermediate domain before training with the real-world off-road driving data.  ...  Fine-tuning the model trained with the synthetic dataset that simulates the off-road driving environment provides more accurate results for the segmentation of real-world off-road driving environments  ...  In [38] , the TL approach is used to semantically segment the off-road scene using the network trained with on-road scenery.  ... 
doi:10.3390/s19112577 fatcat:pnhg2lwnvvbotc5qcdbw6rmezu


Q. Ji, S. Barr, P. James, D. Fairbairn
2018 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences  
Validation of the resulting network is performed using the network layout diagram from the local power company.  ...  In this paper, we develop a geospatial simulation and analysis framework, which is capable of generating fine-scale urban infrastructure networks and storing the network instances in a hybrid database  ...  The synthetic feeders generated by heuristic algorithm are based on ITN road network, which are actual road centrelines.  ... 
doi:10.5194/isprs-archives-xlii-4-291-2018 fatcat:4t5ahntvm5cujcim2upjthudre

Self-Supervised Road Layout Parsing with Graph Auto-Encoding [article]

Chenyang Lu, Gijs Dubbelman
2022 arXiv   pre-print
By using this additional synthetic dataset, which conceptually captures human knowledge of road layouts and makes this available to the network for training, we are able to stabilize and further improve  ...  We create a synthetic dataset containing common road layout patterns and use it for training of the auto-encoder in addition to the real-world Argoverse dataset.  ...  Synthetic road layouts Given the human prior knowledge of road, we also generate a synthetic dataset with each sample representing a random road layout, which is used to train the proposed auto-encoder  ... 
arXiv:2203.11000v2 fatcat:jksy5suirbdgbbpgu5wiuo6kse

Generating All the Roads to Rome: Road Layout Randomization for Improved Road Marking Segmentation [article]

Tom Bruls, Horia Porav, Lars Kunze, Paul Newman
2019 arXiv   pre-print
However, producing the vast quantities of road marking labels required for training state-of-the-art deep networks is costly, time-consuming, and simply infeasible for every domain and condition.  ...  We demonstrate that training on these synthetic pairs improves mIoU of the segmentation of rare road marking classes during real-world deployment in complex urban environments by more than 12 percentage  ...  GENERATING SYNTHETIC TRAINING PAIRS In this section, we explain in detail how to generate synthetic training pairs for road marking segmentation networks to improve performance during real-time deployment  ... 
arXiv:1907.04569v1 fatcat:76nwbyj5fvbp5l3djnzlxgyd6m

London [chapter]

Joan Serras, Melanie Bosredon, Vassilis Zachariadis, Camilo Vargas-Ruiz, Thibaut Dubernet, Mike Batty
2016 The Multi-Agent Transport Simulation MATSim  
Supply The assembly of the supply for our model includes the de nition of the following three components: • road network, • public transport services, and • land-use con guration 1 see  ...  The data used to build the road network is the Integrated Transport Network from the Ordnance Survey.  ...  Figure 73 . 1 : 731 Snapshot of the road network for London's case study colored by road type (le ) and map showing number of bus trips per road segment in London using timetable data (right).  ... 
doi:10.5334/baw.73 fatcat:4fro22pay5elfhctjgdxjwbsqm

Privacy-Preserving Synthetic Location Data in the Real World [article]

Teddy Cunningham, Graham Cormode, Hakan Ferhatosmanoglu
2021 arXiv   pre-print
Our second, network-based approach incorporates public geographic information, such as the road network of a city, to constrain the bounds of synthetic data points and hence improve the accuracy of the  ...  synthetic data.  ...  Road Network Alignment. For Road, we assume that data points are well-aligned with the underlying road network.  ... 
arXiv:2108.02089v1 fatcat:qvymbuhcobhupflob7xxjqdzti

Co-simulation Platform for Developing InfoRich Energy-Efficient Connected and Automated Vehicles [article]

Shunsuke Aoki, Lung En Jan, Junfeng Zhao, Anand Bhat, Ragunathan Rajkumar, Chen-Fang Chang
2020 arXiv   pre-print
To build road networks from the real-world driving data, we develop an Automated Parser and Calculator for Map/Scenario named AutoPASCAL.  ...  Overall, the simulation platform provides a realistic vehicle model, powertrain model, sensor model, traffic model, and road-network model to enable the evaluation of the energy efficiency of eco-autonomous  ...  The updated OpenDRIVE file contains not only the road networks but also the buildings. Also, the OSGB describes the terrain around the road networks.  ... 
arXiv:2004.07980v1 fatcat:fsbabvq2u5ga7ggxx2xr5h3ney

Urban Flood Simulation Using Synthetic Storm Drain Networks

Robert Bertsch, Vassilis Glenis, Chris Kilsby
2017 Water  
Hydrodynamic model results for a synthetically generated and surveyed storm drain inlet network were obtained using the CityCAT 1D/2D system.  ...  the actual network.  ...  Figure 2 . 2 Generation of a synthetic storm drain inlet network: (a) Input data consisting of dissolved road polygon shapefile and pipe network polyline shapefile.  ... 
doi:10.3390/w9120925 fatcat:zzlxkxc3rjamnnjw4kefz2cyuu

ROAD: Reality Oriented Adaptation for Semantic Segmentation of Urban Scenes [article]

Yuhua Chen, Wen Li, Luc Van Gool
2018 arXiv   pre-print
These two modules can be readily integrated with existing state-of-the-art semantic segmentation networks to improve their generalizability when adapting from synthetic to real urban scenes.  ...  Exploiting synthetic data to learn deep models has attracted increasing attention in recent years.  ...  We illustrate the pipeline of our proposed ROAD-net in Fig 1. During training, both synthetic and real images are fed into the network as input.  ... 
arXiv:1711.11556v2 fatcat:kxi3uwenfve4fevcsrsr7yqelq

Real time road signs classification

Paolo Medici, Claudio Caraffi, Elena Cardarelli, Pier Paolo Porta, Guido Ghisio
2008 2008 IEEE International Conference on Vehicular Electronics and Safety  
stretching technique; a multi-layer perceptron neural network is then used to provide a matching score with different road sign models.  ...  This paper describes a method for classifying road signs based on a single color camera mounted on a moving vehicle.  ...  Tests with networks trained on synthetic data are reported in fig. 10 , with one network trained with synthetic data, one network with real signs and the last with synthetic data and real signs.  ... 
doi:10.1109/icves.2008.4640906 dblp:conf/icves/MediciCCPG08 fatcat:njvcmj2zojb6xkwmxbqxkijd5a

Large Scale GPS Trajectory Generation Using Map Based on Two Stage GAN

Xingrui Wang, Xinyu Liu, Ziteng Lu, Hanfang Yang
2021 Journal of Data Science  
Discriminator conditioned on encoded map image restrains generated point sequences in case they deviate from corresponding road networks.  ...  Moreover, our generated trajectories not only indicate the distribution similarity but also show satisfying road network matching accuracy.  ...  Model Road Networks Matching Accuracy Road network matching accuracy is what makes the synthetic trajectory plausible.  ... 
doi:10.6339/21-jds1004 fatcat:dvii7l72uvcyrhcohbji6j3k7y

3D-LaneNet: End-to-End 3D Multiple Lane Detection [article]

Noa Garnett, Rafi Cohen, Tomer Pe'er, Roee Lahav, Dan Levi
2019 arXiv   pre-print
We introduce a network that directly predicts the 3D layout of lanes in a road scene from a single image.  ...  Results are shown on two new 3D lane datasets, a synthetic and a real one.  ...  Other than the aforementioned, the network is identical to the 3D-LaneNet as configured for the synthetic-3D-lanes dataset.  ... 
arXiv:1811.10203v3 fatcat:o2kl63sksjh5taslgfoxw2xumu
« Previous Showing results 1 — 15 out of 48,736 results