Icon

SemanticSegmentation

Semantic Segmentation with Deep Learning in KNIME
Semantic Segmentation with Deep Learning in KNIME This workflow shows how the new KNIME Keras integration can be used to train and deploy a specialized deep neuralnetwork for semantic segmentation.This means that our network decides for each pixel in the input image, what class of object it belongs to.In order to run the example, please make sure you have the following KNIME extensions installed:* KNIME Deep Learning - Keras Integration (Labs)* KNIME Image Processing (Community Contributions Trusted)* KNIME Image Processing - Deep Learning Extension (Community Contributions Trusted)* KNIME Streaming Execution (Beta) (Labs)* KNIME Image Processing - Python Extension (Community Contributions Trusted)You also need a local Python installation that includes Keras. Please refer to https://www.knime.com/deeplearning#keras for installation recommendations and further information. Evaluation Training Deployment PreprocessingWe use only the first 20 of the 183 classes for performance reasons (pixels with a higher class are considered as background). Acknowledgements:The network architecture we use is an adaptation of the U-Net proposed in [1].The dataset we used is taken from [2][1] Ronneberger et al. in "U-Net: Convolutional Networksfor Biomedical Image Segmentation" (https://arxiv.org/abs/1505.04597)[2] Gould et al. "Decomposing a Scene into Geometric andSemantically Consistent Regions." (http://dags.stanford.edu/projects/scenedataset.html) imageslabelingsread labelingsfrom .txt filesonly use first 20classesUNetmodifiedstreaming forperformance andmemory reasonstrain/validation80/20predictedsegmentationtrain for 50 epochs(batch size 4) withAdamground-truthsegmentationread the modelfor deploymentstore fordeployment List Files Image Reader(Table) List Files Python Script (1⇒1) Joiner Extract ID Extract ID DL PythonNetwork Creator Image Preprocessing Partitioning InteractiveSegmentation View Image Viewer Keras NetworkLearner InteractiveSegmentation View StanfordBackground Dataset Fix imagedimensions TensorFlowNetwork Reader Execute and Process Store asTensorFlow Network Semantic Segmentation with Deep Learning in KNIME This workflow shows how the new KNIME Keras integration can be used to train and deploy a specialized deep neuralnetwork for semantic segmentation.This means that our network decides for each pixel in the input image, what class of object it belongs to.In order to run the example, please make sure you have the following KNIME extensions installed:* KNIME Deep Learning - Keras Integration (Labs)* KNIME Image Processing (Community Contributions Trusted)* KNIME Image Processing - Deep Learning Extension (Community Contributions Trusted)* KNIME Streaming Execution (Beta) (Labs)* KNIME Image Processing - Python Extension (Community Contributions Trusted)You also need a local Python installation that includes Keras. Please refer to https://www.knime.com/deeplearning#keras for installation recommendations and further information. Evaluation Training Deployment PreprocessingWe use only the first 20 of the 183 classes for performance reasons (pixels with a higher class are considered as background). Acknowledgements:The network architecture we use is an adaptation of the U-Net proposed in [1].The dataset we used is taken from [2][1] Ronneberger et al. in "U-Net: Convolutional Networksfor Biomedical Image Segmentation" (https://arxiv.org/abs/1505.04597)[2] Gould et al. "Decomposing a Scene into Geometric andSemantically Consistent Regions." (http://dags.stanford.edu/projects/scenedataset.html) imageslabelingsread labelingsfrom .txt filesonly use first 20classesUNetmodifiedstreaming forperformance andmemory reasonstrain/validation80/20predictedsegmentationtrain for 50 epochs(batch size 4) withAdamground-truthsegmentationread the modelfor deploymentstore fordeployment List Files Image Reader(Table) List Files Python Script (1⇒1) Joiner Extract ID Extract ID DL PythonNetwork Creator Image Preprocessing Partitioning InteractiveSegmentation View Image Viewer Keras NetworkLearner InteractiveSegmentation View StanfordBackground Dataset Fix imagedimensions TensorFlowNetwork Reader Execute and Process Store asTensorFlow Network

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: Adrian Nembach
Created at: 2017-11-21
On NodePit since: 2020-05-14
Last update: 2024-04-23
Created with KNIME version: v4.1.2
Tags: deep learningimage processingimage analysiscomputer visionunetneural networksegmentationsemantic segmentationencoder decoderpixel classificationPython

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!