Mastodon

OpenCVThe OpenCV Perception Challenge for Bin-Picking is in full-swing, with teams submitting solutions to the leaderboard. If you're in the robotics or AI space, this is a great way to test your skills, win a share of the $60,000 in prizes, and most excitingly be part of an official CVPR Workshop at CVPR 2025 in Nashville!Sign up to participate: <a href="https://bpc.opencv.org/web/challenges/challenge-page/1/overview" target="_blank" rel="nofollow noopener" translate="no">https://bpc.opencv.org/web/challenges/challenge-page/1/overview</a> See the leaderboard: <a href="https://bpc.opencv.org/web/challenges/challenge-page/1/leaderboard/1" target="_blank" rel="nofollow noopener" translate="no">https://bpc.opencv.org/web/challenges/challenge-page/1/leaderboard/1</a><a href="https://mastodon.social/tags/OpenCV" class="mention hashtag" rel="tag">#OpenCV</a> <a href="https://mastodon.social/tags/ComputerVision" class="mention hashtag" rel="tag">#ComputerVision</a> <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="tag">#AI</a> <a href="https://mastodon.social/tags/BPC2025" class="mention hashtag" rel="tag">#BPC2025</a> <a href="https://mastodon.social/tags/Robotics" class="mention hashtag" rel="tag">#Robotics</a> <a href="https://mastodon.social/tags/Competition" class="mention hashtag" rel="tag">#Competition</a> <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a>

Adam CookOne of the dumbest talks that I have ever heard was given by <a class="mention" href="https://bsky.app/profile/karpathy.bsky.social" rel="nofollow noopener" target="_blank">@karpathy.bsky.social</a> at <a href="https://bsky.app/search?q=%23CVPR" rel="nofollow noopener" target="_blank">#CVPR</a> when he still worked at <a href="https://bsky.app/search?q=%23Tesla" rel="nofollow noopener" target="_blank">#Tesla</a>. An absurd slide deck obviously written by Musk to justify Tesla dumping radar sensors some weeks either - which was only done because of COVID supply chain issues.

José OramasPleased to receive a relatively reduced review assignment (four papers) <a href="https://sigmoid.social/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> 2025. The assignment seems to be on point, thanks for the PCs and ACs for their efforts. <a href="https://sigmoid.social/tags/CV" class="mention hashtag" rel="nofollow noopener" target="_blank">#CV</a>

Adrien FoucartSo apparently <a href="https://social.sciences.re/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> (one of the top computer vision conference) thinks LLM are an acceptable tool to "perform background research" while reviewing a submission... everything is fine, I guess.Also in their rules: "An author cannot submit more than 25 papers." What. Like, I get that a senior author may "significantly contribute" to a handful of papers at the same conf but surely not more than... 5? I can't imagine what they can contribute to 25 -- besides funding of course... <a href="https://social.sciences.re/tags/academia" class="mention hashtag" rel="nofollow noopener" target="_blank">#academia</a>

WaseemHey <a href="https://mastodon.social/tags/fediverse" class="mention hashtag" rel="tag">#fediverse</a>, I want to connect to more people here.I am a <a href="https://mastodon.social/tags/MachineLearning" class="mention hashtag" rel="tag">#MachineLearning</a> engineer who has experience with <a href="https://mastodon.social/tags/python" class="mention hashtag" rel="tag">#python</a>, <a href="https://mastodon.social/tags/golang" class="mention hashtag" rel="tag">#golang</a>, some <a href="https://mastodon.social/tags/rust" class="mention hashtag" rel="tag">#rust</a>, <a href="https://mastodon.social/tags/php" class="mention hashtag" rel="tag">#php</a> and <a href="https://mastodon.social/tags/JavaScript" class="mention hashtag" rel="tag">#JavaScript</a>. I also love research and have a paper published in <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a>. Currently, I am using <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="tag">#AI</a> for improving the safety on the roads. I've worked as full-stack engineer in the past.I find this platform and people here awesome. I've had amazing interactions and want to grow them.<a href="https://mastodon.social/tags/introduction" class="mention hashtag" rel="tag">#introduction</a> <a href="https://mastodon.social/tags/intro" class="mention hashtag" rel="tag">#intro</a> <a href="https://mastodon.social/tags/tech" class="mention hashtag" rel="tag">#tech</a> <a href="https://mastodon.social/tags/mastodon" class="mention hashtag" rel="tag">#mastodon</a> <a href="https://mastodon.social/tags/activitypub" class="mention hashtag" rel="tag">#activitypub</a> <a href="https://mastodon.social/tags/linux" class="mention hashtag" rel="tag">#linux</a>

Oliver GeerI don't think this creative writing is common research practice, but I definitely like it...The paper's called "Eternal Sunshine of the Spotless Net: Selective Forgetting in Deep Networks", and it's from an actual IEEE journal. <a href="https://doi.org/10.1109/CVPR42600.2020.00932" rel="nofollow noopener" translate="no" target="_blank">https://doi.org/10.1109/CVPR42600.2020.00932</a><a href="https://floss.social/tags/academia" class="mention hashtag" rel="nofollow noopener" target="_blank">#academia</a> <a href="https://floss.social/tags/AcademicChatter" class="mention hashtag" rel="nofollow noopener" target="_blank">#AcademicChatter</a> <a href="https://floss.social/tags/computerScience" class="mention hashtag" rel="nofollow noopener" target="_blank">#computerScience</a> <a href="https://floss.social/tags/neuralnetwork" class="mention hashtag" rel="nofollow noopener" target="_blank">#neuralnetwork</a> <a href="https://floss.social/tags/ieee" class="mention hashtag" rel="nofollow noopener" target="_blank">#ieee</a> <a href="https://floss.social/tags/cvpr" class="mention hashtag" rel="nofollow noopener" target="_blank">#cvpr</a> <a href="https://floss.social/tags/cvpr2020" class="mention hashtag" rel="nofollow noopener" target="_blank">#cvpr2020</a> <a href="https://floss.social/tags/machineLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#machineLearning</a> <a href="https://floss.social/tags/machineUnlearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#machineUnlearning</a>

Phil NelsonSome Gameboy Camera pics from CVPR 2024 in Seattle <a href="https://wrestling.social/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> <a href="https://wrestling.social/tags/gameboycamera" class="mention hashtag" rel="nofollow noopener" target="_blank">#gameboycamera</a>

Eryk SalvaggioGreat news: “Because of You” claimed one of 3 top art prizes at the Conference for Computer Vision and Pattern Recognition, <a href="https://assemblag.es/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a>. It’s a video art piece, created with Avijit Ghosh, PhD, exploring parallels between the story of Henrietta Lacks with datafication and the erasure of identity in AI. Thank you to Luba Elliott for recognizing the work — and congrats to the other winners (and creators of all 115 works!). You can read about & see the piece here: <a href="https://thecvf-art.com/project/eryk-salvaggio-avijit-ghosh/" rel="nofollow noopener" target="_blank">https://thecvf-art.com/project/eryk-salvaggio-avijit-ghosh/</a> ➡️

Fraunhofer SITRaphael Antonius Frick präsentierte auf der diesjährigen <a href="https://wisskomm.social/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> 2024 im Rahmen des <a href="https://wisskomm.social/tags/DFAD" class="mention hashtag" rel="nofollow noopener" target="_blank">#DFAD</a> Workshops ein Paper zur Erkennung von Diffusions-basiertem Inpainting. Das Modell kann lokale <a href="https://wisskomm.social/tags/Bildmanipulationen" class="mention hashtag" rel="nofollow noopener" target="_blank">#Bildmanipulationen</a>, die durch Text-to-Image-Generatoren erstellt wurden, mit hoher Präzision identifizieren, selbst wenn diese stark komprimiert wurden oder durch ungesehene Generatoren wie dem kürzlich veröffentlichten <a href="https://wisskomm.social/tags/Adobe" class="mention hashtag" rel="nofollow noopener" target="_blank">#Adobe</a> Firefly 3 erzeugt wurden. Das <a href="https://wisskomm.social/tags/Paper" class="mention hashtag" rel="nofollow noopener" target="_blank">#Paper</a> ist hier zu finden: <a href="https://openaccess.thecvf.com/content/CVPR2024W/DFAD/papers/Frick_DiffSeg_Towards_Detecting_Diffusion-Based_Inpainting_Attacks_Using_Multi-Feature_Segmentation_CVPRW_2024_paper.pdf" rel="nofollow noopener" translate="no" target="_blank">https://openaccess.thecvf.com/content/CVPR2024W/DFAD/papers/Frick_DiffSeg_Towards_Detecting_Diffusion-Based_Inpainting_Attacks_Using_Multi-Feature_Segmentation_CVPRW_2024_paper.pdf</a>

OpenCVOn our next episode we welcome back multi-time guests Paula Ramos and Samet Akcay to tell us what the OpenVINO team have in store for The IEEE/CVF Conference on Computer Vision and Pattern Recognition 2024. <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a> <a href="https://mastodon.social/tags/OpenCV" class="mention hashtag" rel="tag">#OpenCV</a> <a href="https://mastodon.social/tags/OpenVINO" class="mention hashtag" rel="tag">#OpenVINO</a>Watch on LinkedIn: <a href="https://www.linkedin.com/events/7184641695936901120" target="_blank" rel="nofollow noopener" translate="no">https://www.linkedin.com/events/7184641695936901120</a> or YouTube: <a href="https://youtube.com/live/vwUB64GbUNA" target="_blank" rel="nofollow noopener" translate="no">https://youtube.com/live/vwUB64GbUNA</a>

Andrew HundtThanks to the SCoFT team for all the great work that got our paper into <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a> !! 🎉🎊Authors: Zhixuan Liu, Peter Schaldenbrand, Beverley-Claire Okogwu, Wenxuan Peng, Youngsik Yun, Andrew Hundt, Jihie Kim, Jean Oh.Our paper is: SCoFT: Self-Contrastive Fine-Tuning for Equitable Image GenerationThe ArXiV paper link is below with all the details, and we expect to make code available soon!Thanks for reading!<a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a> <a href="https://mastodon.social/tags/CVPR2024" class="mention hashtag" rel="tag">#CVPR2024</a> <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="tag">#AI</a> <a href="https://mastodon.social/tags/ML" class="mention hashtag" rel="tag">#ML</a> <a href="https://mastodon.social/tags/SCoFT" class="mention hashtag" rel="tag">#SCoFT</a><a href="https://arxiv.org/abs/2401.08053" target="_blank" rel="nofollow noopener" translate="no">https://arxiv.org/abs/2401.08053</a>19/n

Andrew HundtSCoFT has been accepted at <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a>! 🎉🎊 Remember Google Gemini’s biased medieval England generated images that were just everywhere? Ancient internet history, I know. I've been chomping at the bit bc we've had methods for more culturally sensitive image generation under review that help to address the kinds of issues we saw with Gemini!SCoFT: Self-Contrastive Fine-Tuning for Equitable Image Generation<a href="https://mastodon.social/tags/CVPR2024" class="mention hashtag" rel="tag">#CVPR2024</a><a href="https://arxiv.org/abs/2401.08053" target="_blank" rel="nofollow noopener" translate="no">https://arxiv.org/abs/2401.08053</a>🧵 1/n

Urs WaldmannSend in your favorite animal papers to our CV4Animals workshop <a href="https://sigmoid.social/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> by March 27!Look forward to the exciting program and hope to see many animal enthusiasts in Seattle <a href="https://sigmoid.social/tags/CVPR2024" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR2024</a>!<a href="https://www.cv4animals.com/" rel="nofollow noopener" translate="no" target="_blank">https://www.cv4animals.com/</a><a href="https://xn--baw-joa.social/@unikonstanz" class="u-url mention" rel="nofollow noopener" target="_blank">@unikonstanz</a> <a href="https://sigmoid.social/tags/CBehav" class="mention hashtag" rel="nofollow noopener" target="_blank">#CBehav</a> <a href="https://sigmoid.social/tags/cv4animals" class="mention hashtag" rel="nofollow noopener" target="_blank">#cv4animals</a> <a href="https://sigmoid.social/tags/computervision" class="mention hashtag" rel="nofollow noopener" target="_blank">#computervision</a>

Urs WaldmannThe 4th CV4Animals workshop will take place at <a href="https://sigmoid.social/tags/CVPR2024" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR2024</a> in Seattle!<a href="https://www.cv4animals.com/" rel="nofollow noopener" translate="no" target="_blank">https://www.cv4animals.com/</a> We invite submissions in 2 tracks: - short 4-page unpublished work (potential invitation to IJCV Special Issue) - published work Deadline: March 27, 2024<a href="https://sigmoid.social/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> <a href="https://sigmoid.social/tags/CVPR24" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR24</a> <a href="https://sigmoid.social/tags/cv4animals" class="mention hashtag" rel="nofollow noopener" target="_blank">#cv4animals</a> <a href="https://xn--baw-joa.social/@unikonstanz" class="u-url mention" rel="nofollow noopener" target="_blank">@unikonstanz</a> <a href="https://sigmoid.social/tags/CBehav" class="mention hashtag" rel="nofollow noopener" target="_blank">#CBehav</a> <a href="https://sigmoid.social/tags/computervision" class="mention hashtag" rel="nofollow noopener" target="_blank">#computervision</a>

Simeon SchmaußThere is also a recording of the presentation at <a href="https://fosstodon.org/tags/CVPR" class="mention hashtag" rel="nofollow noopener" target="_blank">#CVPR</a> <a href="https://youtu.be/9JpGjpITiDM?si=97SN2UCZhMJ1pv_3&t=8377" rel="nofollow noopener" translate="no" target="_blank">https://youtu.be/9JpGjpITiDM?si=97SN2UCZhMJ1pv_3&t=8377</a>

gtbarryPixel Codec AvatarsPiCA improves reconstruction over existing techniques across expressions and views of different gender and skin tone. The PiCA model is much smaller than the state-of-art baseline model, and makes multi-person telecommunication possible: on a single Oculus Quest 2 mobile VR headset, 5 avatars are rendered in realtime in the same scene<a href="https://mastodon.social/tags/facebook" class="mention hashtag" rel="tag">#facebook</a> <a href="https://mastodon.social/tags/meta" class="mention hashtag" rel="tag">#meta</a> <a href="https://mastodon.social/tags/realitylabs" class="mention hashtag" rel="tag">#realitylabs</a> <a href="https://mastodon.social/tags/oculus" class="mention hashtag" rel="tag">#oculus</a> <a href="https://mastodon.social/tags/Pica" class="mention hashtag" rel="tag">#Pica</a> <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a> <a href="https://mastodon.social/tags/avatar" class="mention hashtag" rel="tag">#avatar</a> <a href="https://mastodon.social/tags/communications" class="mention hashtag" rel="tag">#communications</a> <a href="https://mastodon.social/tags/virtualreality" class="mention hashtag" rel="tag">#virtualreality</a> <a href="https://mastodon.social/tags/VR" class="mention hashtag" rel="tag">#VR</a> <a href="https://mastodon.social/tags/technology" class="mention hashtag" rel="tag">#technology</a> <a href="https://mastodon.social/tags/tech" class="mention hashtag" rel="tag">#tech</a><a href="https://research.facebook.com/publications/pixel-codec-avatars/" target="_blank" rel="nofollow noopener" translate="no">https://research.facebook.com/publications/pixel-codec-avatars/</a>

RerunIt’s easy and insightful to use Rerun to understand core concepts of new research! You can find a deep dive into all the papers we have done so far on our new examples page (link below).What papers should we look at next?You can find all of the paper visualizations on our new example page: <a href="https://www.rerun.io/examples/paper-visualizations" target="_blank" rel="nofollow noopener" translate="no">https://www.rerun.io/examples/paper-visualizations</a><a href="https://mastodon.social/tags/computervision" class="mention hashtag" rel="tag">#computervision</a> <a href="https://mastodon.social/tags/ai" class="mention hashtag" rel="tag">#ai</a> <a href="https://mastodon.social/tags/cvpr" class="mention hashtag" rel="tag">#cvpr</a>

LiliWhat is the latest research in quantifying animal movement and behavior?I wrote down a detailed overview here, based on what I saw at CVPR this year: <a href="https://writings.lambdaloop.com/posts/cv4animals-2023/" rel="nofollow noopener" target="_blank">https://writings.lambdaloop.com/posts/cv4animals-2023/</a><a href="https://synapse.cafe/tags/PoseEstimation" class="mention hashtag" rel="nofollow noopener" target="_blank">#PoseEstimation</a> <a href="https://synapse.cafe/tags/animals" class="mention hashtag" rel="nofollow noopener" target="_blank">#animals</a> <a href="https://synapse.cafe/tags/computervision" class="mention hashtag" rel="nofollow noopener" target="_blank">#computervision</a> <a href="https://synapse.cafe/tags/cvpr" class="mention hashtag" rel="nofollow noopener" target="_blank">#cvpr</a> <a href="https://synapse.cafe/tags/cvpr2023" class="mention hashtag" rel="nofollow noopener" target="_blank">#cvpr2023</a> <a href="https://synapse.cafe/tags/neuroscience" class="mention hashtag" rel="nofollow noopener" target="_blank">#neuroscience</a>

Adam CookThe Verge video cites <a href="https://mastodon.social/tags/Musk" class="mention hashtag" rel="tag">#Musk</a>'s desire, in part, in removing certain sensor hardware to cut costs and or to alleviate supply chain issues.**However**, that argument stands in contrast to what then-Autopilot Director, Andrej <a href="https://mastodon.social/tags/Karpathy" class="mention hashtag" rel="tag">#Karpathy</a> stated in the attached <a href="https://mastodon.social/tags/CVPR" class="mention hashtag" rel="tag">#CVPR</a> presentation: <a href="https://www.youtube.com/watch?v=eOL_rCK59ZI&t=28293s" target="_blank" rel="nofollow noopener" translate="no">https://www.youtube.com/watch?v=eOL_rCK59ZI&t=28293s</a>Andrej's argument was, essentially, that "camera-only" based computer vision yielded a "richer dataset" than when Tesla also embraced Radar sensors in their vehicles.

Recent searches

Search options

Administered by:

Server stats:

#cvpr