# Competition end - write up and compare some results

Hi All,
Since this competition has ended, i'm curious to know what routes were taken by you all. I myself had the best results with the Restormer architecture. My write up of my progress can be found here:
[https://toon-beerten.medium.com/denoising-and-reconstructing-dirty-documents-for-optimal-digitalization-ed3a186aa3d6](https://toon-beerten.medium.com/denoising-and-reconstructing-dirty-documents-for-optimal-digitalization-ed3a186aa3d6)
.
We saw that using postprocessing techniques like gaussian blur, the competition score could be improved. But that doesn't necessarily means a better OCR result in the end. That's why i'm very curious to see some actual examples from everybody.
I propose to compare subjectively a few competition images. I chose these four:
0080-lkgp_New-patient-registration-form-280920.pdf-2.png
0132-PAESTA-denying-apollo-moon-landings.pdf-15.png
0346-Erie_sPublicSchools_m_e_ch_20.pdf-16.png
0388-MGMResorts_truth-about-entertainment-infographic.pdf-1.png
Left is original, right is my model output.
![](assets/a519f146c71ed179595b2e1d.png)
![](assets/97b8b6add6295a9fa3173d56.png)
![](assets/4844b8d32e9b18d3ca18c436.png)
![](assets/b14fb0d264438f69dcd34549.png)
---
---
![](assets/860027075bb55accb2a77393.png)
![](assets/8b5498879570990272589b2f.png)
---
---
![](assets/85e35ebf9b95ee1ebfe5865f.png)
![](assets/6d309ba45aa75a5f2bc5ebbd.png)
Hope to see a few results from the others as well!