https://github.com/light4code/tensorflow-research

Research TensorFlow for machine vision industrial usage
https://github.com/light4code/tensorflow-research

anomaly-detection image-processing industrial industrial-automation keras neural-structured-learning nsl segmentation tensorflow unet unet-image-segmentation unet-keras

Last synced: 7 days ago
JSON representation

Research TensorFlow for machine vision industrial usage

• Host: GitHub
• URL: https://github.com/light4code/tensorflow-research
• Owner: Light4Code
• License: mit
• Created: 2020-02-29T12:05:48.000Z (almost 5 years ago)
• Default Branch: master
• Last Pushed: 2022-11-21T22:47:26.000Z (about 2 years ago)
• Last Synced: 2023-03-08T22:56:10.880Z (almost 2 years ago)
• Topics: anomaly-detection, image-processing, industrial, industrial-automation, keras, neural-structured-learning, nsl, segmentation, tensorflow, unet, unet-image-segmentation, unet-keras
• Language: Jupyter Notebook
• Homepage:
• Size: 28.2 MB
• Stars: 5
• Watchers: 6
• Forks: 3
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# tensorflow-research

## Goal

The goal of this repository is to research TensorFlow for machine vision industrial usage. 


It will provide easy to use methods to train models from scratch and also fine tune them afterwards.

## Starting point

To quickly get into the usage you can look into the notebooks, they provide the whole training pipline step by step.

## Backbones

Backbones provide you with a ready to use model architectur.

### Anomaly detection

Anomaly detection will use a autoencoder approche, the prediction substratced from the original image should show the anomaly.

### Segmentation

#### Unet

- [Vanilla Unet](https://arxiv.org/pdf/1505.04597.pdf) (original paper)

- [Custom/Satellite Unet](https://github.com/karolzak/keras-unet)

|           Class            |       Group       |           Backbone            |            Sample notebook            |

|:--------------------------:|:-----------------:|:-----------------------------:|:-------------------------------------:|

|    `ClassificationConv`    |  Classification   |     `classification_conv`     |                                       |

|                            |        ---        |              ---              |                                       |

|     `AutoEncoderConv`      | Anomaly Detection |      `auto_encoder_conv`      | Choco waffle anomaly detection sample |

| `AutoEncoderFullConnected` | Anomaly Detection | `auto_encoder_full_connected` |                                       |

|                            |        ---        |              ---              |                  ---                  |

| `SegmentationVanillaUnet`  |   Segmentation    |  `segmentation_vanilla_unet`  |    Wood plate segmentation sample     |

|            TODO            |   Segmentation    |       `satellite_unet`        |                                       |

|            TODO            |   Segmentation    |         `custom_unet`         |                                       |