项目作者: anega006

项目描述 :
Consists of variety of Autoencoders implementation for various applications such as denoising image, reverse image search, segmantic hair segmentation.
高级语言: Jupyter Notebook
项目地址: git://github.com/anega006/Autoencoders.git
创建时间: 2020-05-19T05:22:35Z
项目社区:https://github.com/anega006/Autoencoders

开源协议:MIT License

下载


Autoencoders

Let us see how the app interface works by inspecting its functionalities.

Here is the main dashboard.

Alt text

A brief description about the project.

Alt text

We clicked “Denoising Dirty Documents” button on the dashboard. Let’s upload an image to see the output.

Alt text

The input image is denoised. We can see the clear output.

Alt text

Here is another input and its clear output.

Alt text

We clicked “Reverse Image Search” button on the dashboard. Let’s upload an image to see the output.

Alt text

The reverse image search produced 5 similar images as output to “Topwear” input image.

Alt text

The reverse image search produced 5 similar images as output to “Watch” input image.

Alt text

The reverse image search produced 5 similar images as output to “Flipflop” input image.

Alt text

The reverse image search produced 5 similar images as output to “Floater” input image.

Alt text

The reverse image search produced 5 similar images as output to “Shoe” input image.

Alt text

The reverse image search produced 5 similar images as output to “Bottomwear” input image.

Alt text

The reverse image search produced 5 similar images as output to “Socks” input image.

Alt text

The reverse image search produced 5 similar images as output to “Bag” input image.

Alt text

The reverse image search produced 5 similar images as output to “Wallet” input image.

Alt text

We clicked “Semantic Hair Segmentation” button on the dashboard. Let’s upload an image to see the output.

Alt text

The output is a predicted binary map and the result applied on actual image.

Alt text

The output is a predicted binary map and the result applied on another actual image.

Alt text