9 Hits in 4.7 sec

Learning to Automatically Catch Potholes in Worldwide Road Scene Images

J. Javier Yebes, David Montero, Ignacio Arriola
2020 IEEE Intelligent Transportation Systems Magazine  
They contained road scenes from different cities in the world, taken with different cameras, vehicles and viewpoints under varied environmental conditions.  ...  Then, we fine-tuned four different object detection models based on Faster R-CNN and SSD deep neural networks.  ...  Grayscale images from a dashcam were used for the purpose and their visual properties were studied.  ... 
doi:10.1109/mits.2019.2926370 fatcat:4e7krkk3gzgftb6aswbvibup5i

Automatic Map Update Using Dashcam Videos [article]

Aziza Zhanabatyrova, Clayton Souza Leite, Yu Xiao
2022 arXiv   pre-print
The algorithm, trained directly with SfM point cloud data, can locate objects detected from 2D images in a 3D space with high accuracy by estimating not only depth from monocular images but also lateral  ...  The errors in the campus area were mainly due to traffic signs seen from a far distance to the vehicle and intended for pedestrians and cyclists only.  ...  Secondly, it provides a novel method for utilizing SfM-based point clouds to train a deep learning model for online pixel-wise 3D localization from monocular RGB data.  ... 
arXiv:2109.12131v2 fatcat:aivdlxkrzfeevk6wo6twz75we4

Driver Behavior Recognition via Interwoven Deep Convolutional Neural Nets with Multi-stream Inputs

Chaoyun Zhang, Rui Li, Woojin Kim, Daesub Yoon, Paul Patras
2020 IEEE Access  
ACKNOWLEDGMENTS This research was supported by grant no. 18TLRP-B131486-02 from the Transportation and Logistics R&D Program funded by the Ministry of Land, Infrastructure and Transport of the Korean government  ...  DEEP LEARNING BASED DRIVER BEHAVIOR IDENTIFICATION Deep learning is becoming increasingly popular for identifying driver behaviors.  ...  Existing research can be categorized into two main classes: non deep learning and deep learning approaches. A.  ... 
doi:10.1109/access.2020.3032344 fatcat:jr2e4ib3lvblhbp3o3pnd3mvue

Zeus: A System Description of the Two-Time Winner of the Collegiate SAE AutoDrive Competition [article]

Keenan Burnett, Jingxing Qian, Xintong Du, Linqiao Liu, David J. Yoon, Tianchang Shen, Susan Sun, Sepehr Samavi, Michael J. Sorocky, Mollie Bianchi, Kaicheng Zhang, Arkady Arkhangorodsky (+5 others)
2020 arXiv   pre-print
With a team of mostly undergraduates and minimal resources, aUToronto has made progress towards a functioning self-driving vehicle, in just two years.  ...  The second year of the competition was held in June 2019 at MCity, a mock town built for self-driving car testing at the University of Michigan.  ...  Members that contributed heavily to the design and development of the Also, a special thanks is given to UofT Engineering for funding UofT's entry in the AutoDrive Challenge.  ... 
arXiv:2004.08752v1 fatcat:jm4os5mzirf7jaogd7a76nrvkq

Exploring Event-driven Dynamic Context for Accident Scene Segmentation [article]

Jiaming Zhang, Kailun Yang, Rainer Stiefelhagen
2021 arXiv   pre-print
Therefore, we propose to extract dynamic context from event-based data with a higher temporal resolution to enhance static RGB images, even for those from traffic accidents with motion blur, collisions  ...  The proposal has been demonstrated to be consistently effective for models learned on multiple source databases including Cityscapes, KITTI-360, BDD, and ApolloScape.  ...  [28] presented a deep reinforcement learning method with anticipation reward and driver fixation reward, which enabled accident prediction on dashcam videos like the DADA dataset [10] .  ... 
arXiv:2112.05006v1 fatcat:xyjr5tiysjgdxbvx2rk2huwdj4

Proceedings_of_Measuring_Behavior_2014.pdf [article]

Andrew Spink, Egon Van Den Broek, Leanne Loijens, Marta Woloszynowska-Fraser, Lucas P. J. J. Noldus
2020 Figshare  
Christian Gutzen and colleagues from Biobserve for thorough hard-and software support. Acknowledgements This work is supported by a European Research Council grant PSARPS.  ...  The services offered are available for a broad range of research projects as well as for external users.  ...  In contrast, model-free approaches estimate a pose directly from observation, without using an accurate 3D model.  ... 
doi:10.6084/m9.figshare.11708187 fatcat:62o74kbgczf2tp2dk64jeszxr4

Scaling data sharing among vehicles to tackle long tail traffic situations

Hongyu Li
Although vehicle-to-vehicle (V2V) communications provide a channel for point cloud data sharing, it is challenging to align point clouds from two vehicles with state-of-the-art techniques due to localization  ...  in the current set of test cases for automated vehicles.  ...  Additional shifts between the cameras and all rotations are simulated by viewpoint transformation of the image from the nearest camera.Then, images are fed into a CNN which then computes a proposed steering  ... 
doi:10.7282/t3-sgra-0w37 fatcat:ixfzljrslzgcbetagzlt4rpeza

Altered Perceptions Discerning the Natural from the Virtual in Extended Reality [article]

Neil Christensen, University Of Calgary, Christian Jacob, Gerald Hushlak
Our view of reality is shaped by our senses, neural processing and learned meanings.  ...  The implementation of applications using elements of photogrammetry, spatial audio and real-time rendering provides a glimpse into present-day capabilities and limitations.  ...  Aligning multiple images, a virtual re-creation of the cameras positions is used to estimate a point cloud and generate 3D geometry.  ... 
doi:10.11575/prism/37321 fatcat:ctbwpadz2jdezetsr6p2aam37y

Reliable scheduling and resource allocation for IoT applications in fog computing

RK Naha
However, managing IoT applications in the Cloud exclusively is not a good solution for some applications, especially for those which are latency-sensitive.  ...  In addition, the user might change their requirements dynamically and also require better reliability from the providers.  ...  The AWS pricing model is the most suitable for pricing estimation in our simulation since we are evaluating our model from a network perspective.  ... 
doi:10.25959/100.00038372 fatcat:26luqbfqrra5dm3mfj4pdmx4wa