Análisis, evaluación e implementación de algoritmos de segmentación semántica para su aplicación en vehículos inteligentes

espanolLos algoritmos de segmentacion semantica, cuyo objetivo es asignar una etiqueta a cada pixel de la imagen, estan adquiriendo una gran relevancia en los ultimos anos. Uno de sus principales ambitos de aplicacion son los sistemas embarcados en vehiculos, donde pueden desempenar distintas funciones para el entendimiento del entorno. Sin embargo, los particulares requisitos de este tipo de aplicaciones, impuestos por las limitadas capacidades de procesamiento disponibles y la complejidad de las escenas, requieren un analisis especifico que vaya mas alla de los parametros clasicos. En este articulo se presenta un analisis detallado de varias arquitecturas contemporaneas para segmentacion semantica en el contexto de su aplicacion en vehiculos, asi como un estudio de su viabilidad en una plataforma real. EnglishSemantic segmentation algorithms, whose goal is to assign a label to each pixel of the image, are acquiring a great relevance in the last years. One of its main areas of application is vehicular embedded, where they can be used to different functions aimed to understand the environment. However, the particular requirements of this type of applications, imposed by the high processing requirements and the complexity of the scenes, require a specific analysis that goes beyond the classical parameters. This article presents a detailed analysis of several contemporary architectures for semantic segmentation in the context of its application in vehicles, as well as a study of its viability in a real platform

[1]  Gang Wang,et al.  DAG-Recurrent Neural Networks for Scene Labeling , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[2]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[3]  Andrew Zisserman,et al.  Very Deep Convolutional Networks for Large-Scale Image Recognition , 2014, ICLR.

[4]  Martin Jägersand,et al.  Deep semantic segmentation for automated driving: Taxonomy, roadmap and challenges , 2017, 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC).

[5]  Luis Miguel Bergasa,et al.  Can we unify monocular detectors for autonomous driving by using the pixel-wise semantic segmentation of CNNs? , 2016, ArXiv.

[6]  Zoubin Ghahramani,et al.  Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference , 2015, ArXiv.

[7]  Iasonas Kokkinos,et al.  DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs , 2016, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Trevor Darrell,et al.  Caffe: Convolutional Architecture for Fast Feature Embedding , 2014, ACM Multimedia.

[9]  Rob Fergus,et al.  Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-scale Convolutional Architecture , 2014, 2015 IEEE International Conference on Computer Vision (ICCV).

[10]  In So Kweon,et al.  VPGNet: Vanishing Point Guided Network for Lane and Road Marking Detection and Recognition , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[11]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[12]  Sebastian Ramos,et al.  The Cityscapes Dataset for Semantic Urban Scene Understanding , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[13]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[14]  Nitish Srivastava,et al.  Improving neural networks by preventing co-adaptation of feature detectors , 2012, ArXiv.

[15]  Eduardo Romera,et al.  ERFNet: Efficient Residual Factorized ConvNet for Real-Time Semantic Segmentation , 2018, IEEE Transactions on Intelligent Transportation Systems.

[16]  Ignacio Parra,et al.  Deep fully convolutional networks with random data augmentation for enhanced generalization in road detection , 2017, 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC).

[17]  Eugenio Culurciello,et al.  ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation , 2016, ArXiv.

[18]  Morgan Quigley,et al.  ROS: an open-source Robot Operating System , 2009, ICRA 2009.

[19]  Torsten Bertram,et al.  A combined recognition and segmentation model for urban traffic scene understanding , 2017, 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC).

[20]  Aurelio Ponz,et al.  IVVI 2.0: An intelligent vehicle based on computational perception , 2014, Expert Syst. Appl..

[21]  José Manuel Pastor,et al.  Visual sign information extraction and identification by deformable models for intelligent vehicles , 2004, IEEE Transactions on Intelligent Transportation Systems.

[22]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[23]  Roberto Cipolla,et al.  Bayesian SegNet: Model Uncertainty in Deep Convolutional Encoder-Decoder Architectures for Scene Understanding , 2015, BMVC.

[24]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[25]  Andreas Geiger,et al.  Augmented Reality meets Deep Learning , 2017, BMVC.