-
Thanks for sharing the source code and the pre-trained model on SIDD validation dataset. I tested the provided pre-trained model on SIDD validation dataset and only got 34.14/0.855 (SPNR/SSIM).
![ima…
-
The code suggests to load the whole training set to memory. There are 2k training images, each image is of size roughly 3000x4000. There are 1 input and 3 gt channels. That makes it:
2000 * 3000 * …
-
## タイトル: StegaVision:注意機構によるステガノグラフィの強化
## リンク: https://arxiv.org/abs/2411.05838
## 概要:
画像ステガノグラフィは、画像内に秘密情報を埋め込む技術です。深層学習の発展により、この分野は大きく進歩しましたが、既存の手法は画質、埋め込み容量、およびセキュリティのバランスに課題がありました。本論文では、エンコーダ…
-
Hi,
I'm looking for an algorithm to normalize images to be able to compare them with QualitySSIM.compute() method.
I've created the below code which works fine but just in case one picture is s…
-
Due to [improvements](https://github.com/sul-dlss/was_robot_suite/issues/462) in wasCrawlDissemination it is now possible to browse Web Archive content using the `webarchive-binary` content-type:
h…
-
안녕하세요. 재애애영 팀의 구자혁입니다.
마지막까지 대회에 힘써주셔서 감사드립니다.
다름이 아니라 방금 전송해주신 최종 공지 내용 관련해서 질문드리려 합니다.
Final evaluation이 리더보드 기준 상위 15~20팀이라고 하셨는데 이 리더보드 기준이라 하면 구글 폼 제출 값을 기준으로 하는 것이 맞을까요??
감사합니다.
-
this project is vary useful,and now i want to know how to use ssim as the loss function for matconvnet. and what PL4NN in the code represent for. thanks very much.
-
안녕하세요!
시간 제한 관련 문의 드립니다.
주어진 코드를 통해 이해한 바로는,
leaderboard public / private 데이터에 대해 reconstruction 하는데 총 3000초가 제한 시간인 것으로 알고 있습니다.
그렇다면,
1. 주어진 vessl gpu에서
2. 최종 output을 뽑는데 있어 현재 주어진 leade…
kjae0 updated
2 months ago
-
Thanks for the code shared, but i have a question: I noticed that the experiment part of your essay include the PSNR, SSIM, LPIPS with their uncertainty. How did you gain those data? I can't find them…
-
I trained the network and ran the code directly where the generator is fpn_inception and generator loss is:
0.5 ∗ Lp + 0.006 ∗ LX + 0.001 ∗ Ladv
And then I trained the network and ran the code wit…