Image Semantic Segmentation based on E-Net with Different Patch-Size Convolution

Computer Vision maybe those fields that aides machine over the understanding of features of images and videos. It is making massive advances in the field of self-driving, robotics, and automation. In the self-driving car, a navigation system must be there to make a decision. The navigation system has to scan the view to decide the next move. By segmenting the image using Semantic Segmentation the navigation system can make a clear decision. Semantic segmentation is the primary step in object detection. In this research paper, various networks do the semantic segmentation are explained. While using different patch size objects can be more utilized. Here in this research demonstrated different patch sizes 3x3, 4x4, 5x5, and 6x6 for training ENET. Deep Neural Network gives the best result in semantic segmentation in terms of Intersection over Union parameter. In the results and analysis section, the results are generated by changing the patch size from the CityScape dataset images. From the analysis, it has been said that 5×5 patch size gives the highest Intersection over Union compared to other deep networks.

[1]  Davide Mazzini,et al.  Guided Upsampling Network for Real-Time Semantic Segmentation , 2018, BMVC.

[2]  Raimondo Schettini,et al.  Deep Multibranch Neural Network for Painting Categorization , 2017, ICIAP.

[3]  Bindhu,et al.  BIOMEDICAL IMAGE ANALYSIS USING SEMANTIC SEGMENTATION , 2019, Journal of Innovative Image Processing.

[4]  Guosheng Lin,et al.  Efficient Piecewise Training of Deep Structured Models for Semantic Segmentation , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Rongrong Ji,et al.  DSNET: Accelerate Indoor Scene Semantic Segmentation , 2019, ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[6]  Sheng Tang,et al.  Scale-Adaptive Convolutions for Scene Parsing , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

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

[8]  Xianxiang Qin,et al.  Image semantic segmentation based on convolutional neural network and conditional random field , 2018, 2018 Tenth International Conference on Advanced Computational Intelligence (ICACI).

[9]  Abul Bashar,et al.  SURVEY ON EVOLVING DEEP LEARNING NEURAL NETWORK ARCHITECTURES , 2019, December 2019.

[10]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[11]  Huimin LU,et al.  Wide Residual Networks for Semantic Segmentation , 2018, 2018 18th International Conference on Control, Automation and Systems (ICCAS).

[12]  George Papandreou,et al.  Rethinking Atrous Convolution for Semantic Image Segmentation , 2017, ArXiv.

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

[14]  Iasonas Kokkinos,et al.  Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs , 2014, ICLR.

[15]  Raimondo Schettini,et al.  A CNN Architecture for Efficient Semantic Segmentation of Street Scenes , 2018, 2018 IEEE 8th International Conference on Consumer Electronics - Berlin (ICCE-Berlin).

[16]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.