Shulou(Shulou.com)06/01 Report--

This article will explain in detail the TensorFlow semantic segmentation suite open source ECCV18 technology BiSeNet real-time segmentation algorithm example analysis, the content of the article is of high quality, so the editor will share it for you to do a reference. I hope you will have a certain understanding of the relevant knowledge after reading this article.

Semantic Segmentation Suite (semantic Segmentation Suite), an open source project on Github, created by George Seif, a machine learning engineer from Building Intelligent Services in the United States, has implemented a large number of the latest semantic segmentation algorithms using Tensorflow. Recently, the open source library has added CVPR2018's newly exposed Dense Decoder Shortcut Connections model and DenseASPP model, and ECCV2018's new real-time semantic segmentation algorithm BiSeNet!

This kind of open source library that follows the cutting edge is exactly what 52CV will vigorously promote!

The goal of Semantic Segmentation Suite is to make it easy for people to code, train, and test the latest semantic segmentation algorithms.

At present, the main functions are:

1) training and testing mode

2) data augmentation

3) contains several latest state-of-the-art semantic segmentation models, and these models can be very easily plug and play.

4) it is convenient to dock any current mainstream semantic segmentation dataset.

5) Evaluation criteria include: precision, recall, F1 score, average accuracy, per-class accuracy, and mean IoU

6) draw the loss function loss and accuracy according to epoch during training

Currently supported feature extraction models:

MobileNetV2, ResNet50/101/152 and InceptionV4.

Currently supported semantic segmentation algorithms:

1) SegNet,arXiv2015

2) SegNet with skip connections,PAMI2017

3) MobileNet-UNet,arXiv2017

4) PSPNet,CVPR2017

5) FC-DenseNet,CVPR2017

6) DeepLabV3,axXiv2017

7) RefineNet,CVPR2017

8) Full-Resolution Residual Networks (FRRN), CVPR2017

9) Global Convolutional Network with Large Kernel,CVPR2017

10) AdapNet,ICRA2017

11) ICNet,ECCV2018

12) DeepLabV3+,ECCV2018

13) DenseASPP,CVPR2018

14) Dense Decoder Shorcut Connections,CVPR2018

15) BiSeNet,ECCV2018

All the state-of-the-art that appeared in the last two years!

The library has built-in sample code for semantic segmentation training, testing, and prediction:

It is convenient to see the result with one click!

The following is an example of the training results on the CamVid dataset using the FC-DenseNet103 model:

On the TensorFlow semantic segmentation suite open source ECCV18 vision technology BiSeNet real-time segmentation algorithm example analysis is shared here, I hope the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.