Full Resolution Residual Networks for Semantic Image Segmentation

Related tags

Deep LearningFRRN
Overview

Full-Resolution Residual Networks (FRRN)

This repository contains code to train and qualitatively evaluate Full-Resolution Residual Networks (FRRNs) as described in

Tobias Pohlen, Alexander Hermans, Markus Mathias, Bastian Leibe: Full Resolution Residual Networks for Semantic Segmentation in Street Scenes. CVPR 2017.

A pre-print of the paper can be found on arXiv: arXiv:1611.08323.

Please cite the work as follows:

@inproceedings{pohlen2017FRRN,
  title={Full-Resolution Residual Networks for Semantic Segmentation in Street Scenes},
  author={Pohlen, Tobias and Hermans, Alexander and Mathias, Markus and Leibe, Bastian},
  booktitle={Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on},
  year={2017}
}

Demo Video

Click here to watch our video.

Installation

Install the following software packages:

  • Python 2.7 or 3.4
  • Numpy
  • Scipy
  • Scikit-Learn
  • OpenCV
  • Theano
    • Scipy
    • Scikit-Learn
  • Lasagne

You may optionally install the following library for better performance.

You can check if all dependencies are installed correctly by running the check_dependencies.py script:

$ python check_dependencies.py --cs_folder=[Your CS folder]
2017-07-26 22:17:34,945 INFO Found supported Python version 3.4.
2017-07-26 22:17:35,122 INFO Successfully imported numpy.
2017-07-26 22:17:35,184 INFO Successfully imported cv2.
2017-07-26 22:17:35,666 INFO Successfully imported sklearn.
2017-07-26 22:17:35,691 INFO Successfully imported sklearn.metrics.
2017-07-26 22:17:35,691 INFO Successfully imported scipy.
Using cuDNN version 6021 on context None
Mapped name None to device cuda: TITAN X (Pascal) (0000:02:00.0)
2017-07-26 22:17:38,760 INFO Successfully imported theano.
2017-07-26 22:17:38,797 INFO Successfully imported lasagne.
2017-07-26 22:17:38,797 INFO Theano float is float32.
2017-07-26 22:17:38,803 INFO cuDNN spatial softmax found.
2017-07-26 22:17:38,807 INFO Use Chianti C++ library.
2017-07-26 22:17:38,826 INFO Found CityScapes training set.
2017-07-26 22:17:38,826 INFO Found CityScapes validation set.

If you don't see any ERROR messages, the software should run on your machine.

Qualitatively evaluation a pre-trained model

Run the script predict.py.

$ python predict.py --help
usage: predict.py [-h] --architecture {frrn_a,frrn_b} --model_file MODEL_FILE
                  --cs_folder CS_FOLDER [--sample_factor SAMPLE_FACTOR]

Shows the predictions of a Full-Resolution Residual Network on the Cityscapes
validation set.

optional arguments:
  -h, --help            show this help message and exit
  --architecture {frrn_a,frrn_b}
                        The network architecture type.
  --model_file MODEL_FILE
                        The model filename. Weights are initialized to the
                        given values if the file exists. Snapshots are stored
                        using a _snapshot_[iteration] post-fix.
  --cs_folder CS_FOLDER
                        The folder that contains the Cityscapes Dataset.
  --sample_factor SAMPLE_FACTOR
                        The sampling factor.

Train a new model

Run the train.py script.

$ python train.py --help
usage: train.py [-h] --architecture {frrn_a,frrn_b,frrn_c} --model_file
                MODEL_FILE --log_file LOG_FILE --cs_folder CS_FOLDER
                [--batch_size BATCH_SIZE]
                [--validation_interval VALIDATION_INTERVAL]
                [--iterator {uniform,weighted}] [--crop_size CROP_SIZE]
                [--learning_rate LEARNING_RATE]
                [--sample_factor SAMPLE_FACTOR]

Trains a Full-Resolution Residual Network on the Cityscapes Dataset.

optional arguments:
  -h, --help            show this help message and exit
  --architecture {frrn_a,frrn_b}
                        The network architecture type.
  --model_file MODEL_FILE
                        The model filename. Weights are initialized to the
                        given values if the file exists. Snapshots are stored
                        using a _snapshot_[iteration] post-fix.
  --log_file LOG_FILE   The log filename. Use log_monitor.py in order to
                        monitor training progress in the terminal.
  --cs_folder CS_FOLDER
                        The folder that contains the Cityscapes Dataset.
  --batch_size BATCH_SIZE
                        The batch size.
  --validation_interval VALIDATION_INTERVAL
                        The validation interval.
  --iterator {uniform,weighted}
                        The dataset iterator type.
  --crop_size CROP_SIZE
                        The size of crops to extract from the full-resolution
                        images. If 0, then now crops will be extracted.
  --learning_rate LEARNING_RATE
                        The learning rate to use.
  --sample_factor SAMPLE_FACTOR
                        The sampling factor.

Monitor training

Start a new notebook server and open training_monitor.ipynb.

License

See LICENSE (MIT).

Copyright

Copyright (c) 2017 Google Inc.

Copyright (c) 2017 Toby Pohlen

Owner
Toby Pohlen
Toby Pohlen
GitHub Repository
This is a yolo3 implemented via tensorflow 2.7

YoloV3 - an object detection algorithm implemented via TF 2.x source code In this article I assume you've already familiar with basic computer vision

2 Jan 17, 2022
AutoPentest-DRL: Automated Penetration Testing Using Deep Reinforcement Learning

AutoPentest-DRL: Automated Penetration Testing Using Deep Reinforcement Learning AutoPentest-DRL is an automated penetration testing framework based o

Cyber Range Organization and Design Chair 217 Jan 01, 2023
DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentation

DCSAU-Net: A Deeper and More Compact Split-Attention U-Net for Medical Image Segmentation By Qing Xu, Wenting Duan and Na He Requirements pytorch==1.1

Qing Xu 20 Dec 09, 2022
Open-source implementation of Google Vizier for hyper parameters tuning

Advisor Introduction Advisor is the hyper parameters tuning system for black box optimization. It is the open-source implementation of Google Vizier w

tobe 1.5k Jan 04, 2023
RCT-ART is an NLP pipeline built with spaCy for converting clinical trial result sentences into tables through jointly extracting intervention, outcome and outcome measure entities and their relations.

Randomised controlled trial abstract result tabulator RCT-ART is an NLP pipeline built with spaCy for converting clinical trial result sentences into

2 Sep 16, 2022
Implementation of ProteinBERT in Pytorch

ProteinBERT - Pytorch (wip) Implementation of ProteinBERT in Pytorch. Original Repository Install $ pip install protein-bert-pytorch Usage import torc

Phil Wang 92 Dec 25, 2022
Semantic code search implementation using Tensorflow framework and the source code data from the CodeSearchNet project

Semantic Code Search Semantic code search implementation using Tensorflow framework and the source code data from the CodeSearchNet project. The model

Chen Wu 24 Nov 29, 2022
The code used for the free [email protected] Webinar series on Reinforcement Learning in Finance

Reinforcement Learning in Finance [email protected] Webinar This repository provides the code f

Yves Hilpisch 62 Dec 22, 2022
PyTorch implementation for Graph Contrastive Learning with Augmentations

Graph Contrastive Learning with Augmentations PyTorch implementation for Graph Contrastive Learning with Augmentations [poster] [appendix] Yuning You*

Shen Lab at Texas A&M University 382 Dec 15, 2022
Jremesh-tools - Blender addon for quad remeshing

JRemesh Tools Blender 2.8 - 3.x addon for quad remeshing. Currently it is a wrap

Jayanam 89 Dec 30, 2022
The tool under this branch fork can be used to crack devices above A12 and up to A15. After cracking, you can also use SSH channel strong opening tool to open SSH channel and activate it with Demo or Shell script. The file can be extracted from my Github homepage, and the SSH channel opening tool can be extracted from Dr238 account.

Welcome to C0xy-A12-A15-Attack-Tool The tool under this branch fork can be used to crack devices above A12 and up to A15. After cracking, you can also

13 Dec 23, 2022
PanopticBEV - Bird's-Eye-View Panoptic Segmentation Using Monocular Frontal View Images

Bird's-Eye-View Panoptic Segmentation Using Monocular Frontal View Images This r

63 Dec 16, 2022
Official Code Release for Container : Context Aggregation Network

Container: Context Aggregation Network Official Code Release for Container : Context Aggregation Network Comparion between CNN, MLP-Mixer and Transfor

peng gao 42 Nov 17, 2021
Use unsupervised and supervised learning to predict stocks

AIAlpha: Multilayer neural network architecture for stock return prediction This project is meant to be an advanced implementation of stacked neural n

Vivek Palaniappan 1.5k Jan 06, 2023
Hitters Linear Regression - Hitters Linear Regression With Python

Hitters_Linear_Regression Kullanacağımız veri seti Carnegie Mellon Üniversitesi'

AyseBuyukcelik 2 Jan 26, 2022
A stable algorithm for GAN training

DRAGAN (Deep Regret Analytic Generative Adversarial Networks) Link to our paper - https://arxiv.org/abs/1705.07215 Pytorch implementation (thanks!) -

195 Oct 10, 2022
CVPRW 2021: How to calibrate your event camera

E2Calib: How to Calibrate Your Event Camera This repository contains code that implements video reconstruction from event data for calibration as desc

Robotics and Perception Group 104 Nov 16, 2022
METER: Multimodal End-to-end TransformER

METER Code and pre-trained models will be publicized soon. Citation @article{dou2021meter, title={An Empirical Study of Training End-to-End Vision-a

Zi-Yi Dou 257 Jan 06, 2023
Train Scene Graph Generation for Visual Genome and GQA in PyTorch >= 1.2 with improved zero and few-shot generalization.

Scene Graph Generation Object Detections Ground truth Scene Graph Generated Scene Graph In this visualization, woman sitting on rock is a zero-shot tr

Boris Knyazev 93 Dec 28, 2022
Selene is a Python library and command line interface for training deep neural networks from biological sequence data such as genomes.

Selene is a Python library and command line interface for training deep neural networks from biological sequence data such as genomes.

Troyanskaya Laboratory 323 Jan 01, 2023