Scott Wisdom

Veo 3 is here, and in addition to better visuals, it makes noises and speaks! This was a massive effort made possible by incredible passion from the whole Veo team and the many other team enabling it to launch today. Looking forward to seeing what others do with it! #veo3

Veo 3, our SOTA video generation model, has native audio generation and is absolutely mindblowing. For filmmakers + creatives, we’re combining the best of Veo, Imagen and Gemini into a new filmmaking tool called Flow. Ready today for Google AI Pro and Ultra plan subscribers.

We're sharing progress on our video-to-audio (V2A) generative technology. 🎥 It can add sound to silent clips that match the acoustics of the scene, accompany on-screen action, and more. Here are 4 examples - turn your sound on. 🧵🔊 dpmd.ai/v2a

It's so awesome to see the impact of the computational audio capabilities we developed featured in @madebygoogle 🎉 🎉 🎉 Congrats to John Hershey, @ScottTWisdom, @PGetreuer & everyone who contributed for pioneering new computational audio capabilities in Pixel8 #MadeByGoogle

Sorry it took forever (I did the editing this year...): videos of all #SANE2022 talks by @TweetRupal @mhnt1580 @ScottTWisdom @tnsainath @shinjiw_at_cmu @anoopcherian @gan_chuang are finally available! Here's the essential binge-watching YouTube playlist👇 youtube.com/playlist?list=…

Strong showing at #SANE2022 to learn about the latest and greatest in speech and audio research from a stellar lineup!

Here is a short presentation of AudioScopeV2!📢 @ScottTWisdom and I are looking forward to discussing further about open-domain on-screen sound separation and meeting you in #ECCV2022! webpage:google-research.github.io/sou... arxiv:arxiv.org/abs/2207.10141 video:youtu.be/6UgcS3NdPn8

Full list of speakers and talk details for #SANE2022 (Thursday 10/6, Cambridge, MA) now available! @anoopcherian @gan_chuang @mhnt1580 @TweetRupal @tnsainath @shinjiw_at_cmu @ScottTWisdom Poster & demo submissions due 9/21. Registration/Details: saneworkshop.org

I am 😃 that we will present AudioScopeV2 at #ECCV2022! If you want to learn about improved audio-visual attention models and calibration for on-screen sound separation check our paper w. @ScottTWisdom! project-page: google-research.github.io/sound-separati… new dataset: github.com/google-researc…

Distance-Based Sound Separation abs: arxiv.org/abs/2207.00562 project page: google-research.github.io/sound-separati… With a single nearby speaker and four distant speakers, the model improves scale-invariant signal to noise ratio by 4.4 dB for near sounds and 6.8 dB for far sounds

SANE is back! Thursday, Oct. 6 in Kendall Square, Cambridge, MA. Confirmed speakers: A. Cherian @anoopcherian, C. Gan @gan_chuang, W.-N. Hsu @mhnt1580, T. Sainath @tnsainath, S. Watanabe @shinjiw_at_cmu, S. Wisdom @ScottTWisdom. More details: saneworkshop.org

Happy to see my summer work with @ScottTWisdom, Hakan Erdogan, and John Hershey was accepted for presentation at @ieeeICASSP 2022 😊 My first ICASSP paper in the books! Immensely thankful for their mentorship. Our first version can be found on arXiv at: arxiv.org/abs/2110.10739

We can learn a lot about our environment just by listening to the birds. New #GoogleAI approaches can help isolate and identify birdsongs, helping ecologists better understand food systems and forest health. 🐦 ai.googleblog.com/2022/01/separa…

Our paper received a #WASPAA2021 special award for *Best Audio Representation Learning Paper*: "Self-Supervised Learning from Automatically Separated Sound Scenes". 🎉🚀 paper: arxiv.org/abs/2105.02132 talk: youtu.be/Tts5vYmGwUY slides: bit.ly/3lBjAnr 👇

Our DF-Conformer paper has received the “Best Speech Enhancement Paper Award” from #WASPAA2021! Yay!!

🔊Here's the video presentation of our WASPAA21 paper: "Self-Supervised Learning from Automatically Separated Sound Scenes". Work done during an internship at Google Research. paper: arxiv.org/abs/2105.02132 video: youtu.be/Tts5vYmGwUY slides: bit.ly/3lBjAnr

🔊Happy to announce FSD50K: the new open dataset of human-labeled sound events! Over 51k Freesound audio clips, totalling over 100h of audio manually labeled using 200 classes drawn from the AudioSet Ontology. Paper: arxiv.org/pdf/2010.00475… Dataset: doi.org/10.5281/zenodo…

I am thrilled to announce that our paper "Unsupervised Sound Separation using Mixtures of Mixtures" got accepted to #NeurIPS2020 as a #Spotlight paper!! 📢📢 All kudos to @ScottTWisdom and the rest of the Google guys! arxiv.org/pdf/2006.12701…

We are very happy to announce that all the videos of our recent #ICML2020 workshop on self-supervised learning are now publicly available at slideslive.com/icml-2020/self… Thanks #ICML2020 and @SlidesLive for that! @MILAMontreal #DeepLearning #AI #Speech #MachineLearning

Glad you like it, thanks for the nice summary!