SVRHM 2020 - Concluding Remarks
Date Posted:
December 16, 2020
Date Recorded:
December 12, 2020
All Captioned Videos SVRHM Workshop 2020
Description:
Diversity in AI Best Paper Award (NVIDIA Titan RTX) Ceremony and Oculus Quest Award for Breakthrough in Biologically inspired Generative Models.
ARTURO DEZA: So again, just thanking all the sponsors and you guys for being here, sticking around for the 8, 10, 12 minutes, or hours, or seconds, depending on the time that we had. But really, thank you all for being here.
For those who weren't here in the morning, I just want to do a quick recap of some fun facts about SVRHM. It was the first iteration, even though it's the second edition, where we had invited talks for orals. Robert Geirhos, Aviv Netanyahu, Salman Khan, and Melanie Sclar gave invited talks and presentations that will be recorded and up online.
Another fun fact, for those who weren't here in the morning, there's pretty much no virtual difference in terms of gender at this point for the highest scoring papers. And two of the invited talks were orals purely, just on the rating were by two women who are first authors from different institutions. So this is very exciting and great.
Another fun fact, there were two authors with first single author papers actually, once again from different parts of the world as well, Germany and India. They had their [INAUDIBLE] earlier. And I'm sure that if you reach out to them, they're more than happy to talk to you about their work.
There was one paper by an undergraduate, a student who does not have a bachelor's degree yet, a student of yours truly and Tomaso Poggio here at MIT. Again, feel free to reach out to him if you have any questions.
There were other two papers of undergraduates. So again, still students who don't have a bachelor's yet, but are in the process of doing it, from also the previous mentioned paper by the Universidad de Buenos Aires. This is Sebastian Bujia, and two other students from Turkey, again, from all over the world, Muhammed Samil Atesoglu and Cagatay Yigit.
Two accepted papers with a master's students as a first-author. Interestingly, again, a single author paper, Sashi Kant Gupta, and Johannes Singer.
Other things that we talked about earlier, in terms of the-- again, for those who missed it, I talked about it earlier, but just doing a quick recap, papers that were submitted and published in SVRHM last year, where they ended up today. This paper was nominated for the Best Paper Award. It had an honorable mention. And it ended up being published at Neural Computation. So I encourage all of you to see it if you log in to the website. It's a paper on color and spatial poignancy by Ethan Harris, Daniela Mihai, and Jonathan Hare.
The very famous follow-up paper on the Geirhos [INAUDIBLE] paper on texture bias, but this one that's created a buzz in NeurIPS this year, "Origin and Prevalence of Texture Bias." Actually, its preceding paper won the Best Poster Award last year at SVRHM. For those of who were there, this was a work by Katherine Hermann at the time only with Simon Kornblith, and Ting Chen later is in this-- in the NeurIPS edition in this main conference for this paper. And I believe that is that.
There's a couple of more slides, but I just kind of want to cut to the chase-- give away the awards before we wrap up the session. After internal discussions with all the organizers, the Facebook Reality Labs Breakthrough in Biologically-Inspired Models Best Paper Award goes to-- drum roll-- this is the group at MIT with first author of Aviv Netanyahu, Tianmin Shu, Boris Katz, Andrei Barbu, and Josh Tenenbaum. For across the three reviewers, they all gave this paper a 9. It was very new.
Aviv gave a talk earlier today. And for those who missed it, the talk was online. So I encourage you all to look at it. Again, the paper and the scores are available on OpenReview. So you're all welcome to look at that. We'll coordinate as well. Feel free-- Aviv or anyone else who's here and very excited about the award, just contact any of the organizers about this.
And the NVIDIA Diversity in AI Best Paper Award, actually, happened to be the second highest scoring paper with scores 7, 8, and 9, also available on OpenReview, is by Melanie Sclar, Gaston Bujia, Sebastian Vita, Guillermo Solovey, and Juan Kamienkowski from Buenos Aires, Argentine, from different institutions, Laboratorio de Inteligencia Artificial Aplicada, Universidad de Buenos Aires, and Instituto del Cálculo. And so congratulations to you guys as well.
Again, I don't know how many of the speakers before are here. But we do give someone an Oscar, the award. And that's great. Oh, [INAUDIBLE] has raised hand.
MELANIE SCLAR: Is Juan also available? I'm Melanie, actually, but I'm logged in from the Society of Neuroscience Zoom.
ARTURO DEZA: Cool.
MELANIE SCLAR: But thank you so much. We were so happy for the award.
I don't know if anyone else is-- any other author is around, but we are very happy. Yeah! Thank you so much.
ARTURO DEZA: Cool. Great. Great. Yeah. Thank you guys for submitting as well, because it's also hard to get people sometimes interested. So thank you guys.
I see Andrei, Andrei Barbu, maybe you want to-- who is an author, maybe you can--
ANDREI BARBU: Hi, folks. Yeah. Thank you so much. That's really amazing. Definitely, if you're interested, check out check out the data set in the simulator. I think this is just the first step in getting social interactions, and physics, and vision all together in one package. That's really going to be necessary for any kind of robots that we have that interact with us. So thanks a lot. We're very honored.
ARTURO DEZA: Cool. Great. Can you-- I think, yeah, you guys can hear me. So thank you everybody, really. And on that note, I'd like to thank all the speakers who are still here, who were here in this [INAUDIBLE] session, all the sessions, all the attendees, really. So I just kind of give them a big round of applause.
And also-- well, from the organizing committee, to all of you for being here, and making this workshop a success in a way. And this is the second time we're doing it. We hope that there's a third time. And again, in informal conversations with many friends and colleagues, we've always just been saying that one day we're kind of dreaming of this to be like maybe its own NeurIPS or ICLR maybe one day. And that would be amazing.
So yeah, that's pretty much it. I don't know, Ratan, if you have any other words as well.
APURVA RATAN MURTY: Just thanks to everyone who has stayed on with us, [INAUDIBLE] many parts of the world. And if you have any feedback, please do let us know. And yeah, thanks again to everyone.
ARTURO DEZA: Cool. All right. Goodbye, everybody. Good night, and thank you. Till next year.
APURVA RATAN MURTY: Yep.
ARTURO DEZA: Bye.
APURVA RATAN MURTY: Until next year.
have an interactive transcript feature enabled, which appears below the video when playing. Viewers can search for keywords in the video or click on any word in the transcript to jump to that point in the video. When searching, a dark bar with white vertical lines appears below the video frame. Each white line is an occurrence of the searched term and can be clicked on to jump to that spot in the video.
