Building layout tagging

This deep learning model tags the images of building layout, like

Input	Output
	{ 'tag': 'building_layout', 'score': 0.99993956 }
	{ 'tag': 'non_building_layout', 'score': 0.6236702 }

Installation

to install the building layout tagger, you need Python 3.7.7

git clone https://github.com/liang6261515/building_layout_tagging.git

cd building_layout_tagging

pip3 install -r requirements.txt

wget https://github.com/fchollet/deep-learning-models/releases/download/v0.4/xception_weights_tf_dim_ordering_tf_kernels_notop.h5

download the pretrain model of cheque detection from this url https://drive.google.com/file/d/1JPrSCBMQAzmY8rSvNgMt76fbv2tpuga3/view?usp=sharing to the folder building_layout_tagging

Using

Test case 1

let's download a building layout image

wget https://www.crismatec.com/python/of/office-building-floor-plans-beautiful-design-layout-plan_office-decoration.jpg

this image looks like

then open you python3, and import the tagger

from building_layout_tagging import building_layout_tagging

and run the tagging program

building_layout_tagging('office-building-floor-plans-beautiful-design-layout-plan_office-decoration.jpg')

you will the output:

{'tag': 'building_layout', 'score': 0.99993956}

Test case 2

let's dowload another image which looks like building layout but actually it is not

wget https://digitalcommunications.wp.st-andrews.ac.uk/files/2017/02/track-changes-example-1.png

this image looks like

run my tagger against it

building_layout_tagging('track-changes-example-1.png')

and the program says

{'tag': 'non_building_layout', 'score': 0.6236702}

I am looking for a job

feel free to contact me if you have any problem with this package or you are hiring data scientist/AI engineer. I am actively looking for data science/AI related jobs

My email: yanliang2345@outlook.com