Papers of the day   All papers

Robust Image Colorization using Self Attention based Progressive Generative Adversarial Network


Aug 31 2019 Jason Antic

I just saw this paper. It's literally DeOldify. Yet no mention. I know this likeness is not a coincidence because I talked to one of the authors about it months ago before it went up. I won't name names.
56 replies, 516 likes

Aug 31 2019 Jeremy Howard

This is some pretty shitty behavior. All the work put into this and now some folks come and rip it off in its entirety.
9 replies, 265 likes

Aug 31 2019 Rogue 🌻. Bigham

imagine if all computer scientists had even a basic training in HCI, maybe they’d do a tiny bit of thinking before doing random ML that is at best pointless and possibly negative impact on the world
8 replies, 57 likes

Aug 31 2019 Sanyam Bhutani

Hello Twitter World, Requesting you all to help here: This is really messed up: @citnaj (Jason Antic)'s truly truly amazing work (DeOldify Project) has been copied (stolen) and claimed by the authors as their own in a published paper. OP by JasonπŸ‘‡
4 replies, 49 likes

Aug 31 2019 abhishek thakur

Plagiarism once again!!!!. Plagiarism at any level is disgusting, whether it may be a simple blog post or a research paper. People have to learn to give proper credits!
0 replies, 24 likes

Sep 01 2019 François Fleuret

This thread to remember that--all ethical aspects aside--although plagiarism may sometime look like a good idea it is definitely not a safe strategy. This is what the negative part of the reward looks like.
0 replies, 20 likes

Aug 31 2019 Aakash Kumar Nain πŸ”Ž

Cc: @abhi1thakur @bhutanisanyam1 Ooo the IITians.....again!! πŸ˜‚πŸ˜‚
3 replies, 13 likes

Aug 31 2019 Uri Manor

Fellow academics - what is the best way to handle this situation?
5 replies, 11 likes

Aug 31 2019 Delip Rao

This is just awful. Jason has spent a ton of time creating and refining DeOldify. Things like this undermine people’s generosity. Hey @IITJodhpur @CSIR_IND, look at the shitty stuff your students are doing. You are better than this.
0 replies, 7 likes

Sep 03 2019 HotComputerScience

Most popular computer science paper of the day: "Robust Image Colorization using Self Attention based Progressive Generative Adversarial Network"
0 replies, 5 likes

Aug 31 2019 Pablo Munoz

Please, don’t steal qork from other people and then apologize when you are caught
0 replies, 4 likes

Aug 31 2019 Graham Dellaire

Given @jwoodgett G&M article and discussion of research ethics, here is a shocking example 🀯 of breach of ethics regarding blatant plagiarism, failure to cite work and provide authorship - every collaborative scientists nightmare 😳😱πŸ₯΅ @cyclocrossguy @LandonGetz @mari_kaartinen
1 replies, 4 likes

Aug 31 2019 Julian Harris

DeOldify is possibly the most innovative progress in automated photo & film colourisation for a long time, if not ever. Jason has been exceptionally generous in open sourcing his work and been clear that a paper was in the works. But someone else decided to publish it first.
0 replies, 3 likes

Aug 31 2019 πŸŽ™Jean-Louis Queguiner

When an inventor that provide open source stuffs,agree to spend time with others to help them and share his research and than those same people backfire at him... this is all what discussed me in research ... i am a bit extremist but in school it’s called cheating and ...
1 replies, 3 likes

Sep 01 2019 Rana Hanocka

This type of blatant plagerism is completely unacceptable. I hope proper action will be taken by @CVPR @cvpr2019
0 replies, 2 likes

Aug 31 2019 Ashwin Saxena

IIT Jodhpur student @Megh81807309 and team copied @citnaj's work without ANY credit for CVPR. Very shitty thing to copy someone else's research and pass it off as your own.
0 replies, 1 likes

Aug 31 2019 Oliver Zettinig

Good example how *not* to do science. Behavior of these copycat paper authors, including in this thread, is pathetic. They should all be banned for a few years from #CVPR.
0 replies, 1 likes

Aug 31 2019 πš„πš•πšžπšπš‹πšŽπš” πš‚. π™ΊπšŠπš–πš’πš•πš˜πšŸ

This thread is fun to read. It also includes a fascinating response by the authors of the paper.
0 replies, 1 likes