IS之CNN之SegNet：基于深度学习的图像分割SegNet算法的简介、关键思路、使用方法、案例应用之详细攻略

IS之CNN之SegNet：基于深度学习的图像分割SegNet算法的简介、关键思路、使用方法、案例应用之详细攻略

导读
基于CNN的神经网络SegNet算法可进行高精度地识别行驶环境。

 

SegNet算法的简介

论文《SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation》
arXiv地址：https://arxiv.org/abs/1511.00561?context=cs
PDF地址：https://arxiv.org/pdf/1511.00561.pdf

Vijay Badrinarayanan, Kendall, and Roberto Cipolla（2015）: SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. arXiv preprint arXiv:1511.00561 (2015).

 

SegNet算法的关键思路

1、An illustration of the SegNet architecture. There are no fully connected layers and hence it is only convolutional. A decoder upsamples its input using the transferred pool indices from its encoder to produce a sparse feature map(s). It then performs convolution with a trainable filter bank to densify the feature map. The final decoder output feature maps are fed to a soft-max classifier for pixel-wise classification.

2、An illustration of SegNet and FCN [2] decoders. a, b, c, d correspond to values in a feature map. SegNet uses the max pooling indices to upsample (without learning) the feature map(s) and convolves with a trainable decoder filter bank. FCN upsamples by learning to deconvolve the input feature map and adds the corresponding encoder feature map to produce the decoder output. This feature map is the output of the max-pooling layer (includes sub-sampling) in the corresponding encoder. Note that there are no trainable decoder filters in FCN.

 

 

SegNet算法的使用方法

1、Results on CamVid day and dusk test samples. SegNet shows superior performance, particularly with its ability to delineate boundaries, as compared to some of the larger models when all are trained in a controlled setting. DeepLab-LargeFOV is the most efficient model and with CRF post-processing can produce competitive results although smaller classes are lost. FCN with learnt deconvolution is clearly better. DeconvNet is the largest model with the longest training time, but its predictions loose small classes. Note that these results correspond to the model corresponding to the highest mIoU accuracy in Table 3.

2、Qualitative assessment of SegNet predictions on RGB indoor test scenes from the recently released SUN RGB-D dataset [23]. In this hard challenge, SegNet predictions delineate inter class boundaries well for object classes in a variety of scenes and their view-points. Overall rhe segmentation quality is better when object classes are reasonably sized but is very noisy when the scene is more cluttered. Note that often parts of an image of a scene do not have ground truth labels and these are shown in black colour. These parts are not masked in the corresponding deep model predictions that are shown. Note that these results correspond to the model corresponding to the highest mIoU accuracy in Table 4.

 

 

 

SegNet算法的案例应用

1、Fig. 1. SegNet predictions on road scenes and indoor scenes. To try our system yourself, please see our online web demo at http://mi.eng.cam.ac. uk/projects/segnet/