FaRo

Happy to see that DeMask built with asteroid won the first place of the @PyTorch summer #hackathon! The model allows you to enhance muffled speech when wearing facemask Demo: youtu.be/QLf10Uqu8Yk 👋 to our team: @mnlpariente @michelolzam @_jonashaag and Samuel Cornell

DJs making music magic, create yours with AudioShake stems. djay Pro from @algoriddim now includes AudioShake’s tech to isolate vocals, instruments, and drums at the highest quality available to the industry. On mobile. 🎧📷#djtools #beatmaking

Our paper on lyrics transcription evaluation is on arXiv with updated and extended results (including Whisper v3)! 📄 arxiv.org/abs/2311.13987 Also, the benchmark is now on @paperswithcode: 🏆 paperswithcode.com/dataset/jam-alt 👏 @faroit @h_schreiber @Luke_Miner @cnst_ant @AudioShakeAI

We just released Jam-ALT, a formatting- and punctuation-aware automatic lyrics transcription benchmark (based on the JamendoLyrics dataset) that follows music industry guidelines. 🧵 🔎 audioshake.github.io/jam-alt/ 🤗 huggingface.co/datasets/audio… 🧑‍💻 github.com/audioshake/alt…

@ISMIRConf its 3PM and its not open :-/

📢 LATE-BREAKING DEMO REOPENS 📢 today at 3 p.m. (CEST) and will remain open until the originally announced deadline, We can guarantee acceptance of a limited number of papers (venue capacity will be confirmed later) and apply a priority to papers submitted early. #ISMIR2023

@zhaojw1998 @gkspearow closing this after a couple of hours doesn't sound like a good review process to me

@gkspearow I believe that is the most likely reason. Since there are only 15 entries to be accepted, I guess it's reasonable that all spots are filled quite early.

ISMIR LBD just started and acceptance is rolling-based. CMT still open this morning but closed just now before I finished my draft. Never expected such competitiveness🥲 But anyw, seems many people have great progress to share. Look forward to connect with #ISMIR2023 in Milan :-)

@zhaojw1998 same here for us. Very unfortunate and not really fair as it can't be automated. Why not let more papers to be submitted and review them by quality instead of submission time?

@serrjoa Btw. Awesome paper!

Want to convert from mono to stereo? 🔊➡️🔊🔊 In our latest work, we posit that upmixing mono to stereo is a great avenue for generative modeling, and that parametric stereo coding can facilitate things. Paper: arxiv.org/abs/2306.14647

@serrjoa Out of curiosity, is that parametric stereo robust enough to extracting spatial parameters from a mixture and apply it on sources?

@JOSS_TheOJ The main author is github.com/neillu23 and paper submission can be found here: github.com/openjournals/j…

I am looking for reviewers for a @JOSS_TheOJ submission with expertise in speech and python. The software under review is a new speech enhancement module of github.com/espnet/espnet (hence its so difficult to get reviewers without conflicts of interest). Any pointer is helpful!

@ParcolletT @julien_c Montpellier!

Hugging face Lyon office

Join us for a special edition of our Mila Music + AI Reading Group from February 8th to 22nd! We're excited to host 5 teams from the 2022 AI Song Contest, an international contest where musicians and scientists collaborate to explore human-ai co-creativity.

#Pytorch implementation of MusicLM, new SOTA model for music generation using attention networks plus embeddings from MuLan, a text-audio contrastive learned model github.com/lucidrains/mus…

@DrJimFan @WilliamLamkin Reason for that are conference deadlines though

@WilliamLamkin Same, I’m surprised by the momentum in audio this year. 4 models in one week is insane even by modern AI’s pace.

Music & sound effect industry has not fully understood the size of the storm about to hit. There’re not just one, or two, but FOUR audio models in the past week *alone* If 2022 is the year of pixels for generative AI, then 2023 is the year of sound waves. Deep dive with me: 🧵

🎻 The SDX23 challenge introduces a new formulation of audio source separation: cinematic sound separation. The task is to separate a movie's audio into three tracks: dialogue, sound effects & music. 📕 Give it a try using the starter kit. aicrowd.com/challenges/sou…

#AudioLDM, the text-to-audio model, is now available on HuggingFace and GitHub to play with! We will add more functionality and further improve the model performance in the near future. Share the interesting samples you generate! github.com/haoheliu/Audio… huggingface.co/spaces/haoheli…

@naotokui_en @csteinmetz1 Training is just a very small aspect of it. Lawyers will first go after the obvious things: the startups that can generate new Taylor Swift songs.

@csteinmetz1 It can be problematic because it’s completely legal to train AI models on copyrighted material in some countries (including Japan)

No one would dare train their music generation diffusion model on copyrighted music so we have nothing to worry about... 🫠

I guess developers are thrilled!

@JonathanLeRoux @ethanmanilow @r4b1tt ^

@JonathanLeRoux @ethanmanilow For the challenge we run mean across songs. But many papers and also (rightfully) show median across songs as outliers can be dramatic.

Hey music separation researchers. We added a new definition of the SDR metric when we launched the last @sounddemix. To make it less confusing for future papers, we want to rename the metric. Please vote