IT'S FINALLY OUT! After months of work, I'm happy to announce that the first video in a series on visualizing deep learning is out. It's on visualizing the structure of a neural network. RTs appreciated :) youtu.be/UOvPeC8WOt8

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: Manuel Pariente Michel Olvera @_jonashaag and Samuel Cornell

Woohoo 🤩 We won the @Pytorch #hackathon with DeMask - built with Asteroid ! Truly surprised and humbled, but also very proud of our team (Samuele Cornell, FaRo, Michel Olvera and @_jonashaag). Project page: devpost.com/software/aster… Demo: youtube.com/watch?v=QLf10U… Thanks !

🔊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…

Día de los Muertos, or #DayoftheDead, unfolds in an explosion of color and joy demonstrating love and respect for lost loved ones on.natgeo.com/3kPrIh9

Interested in machine listening? Curious how robots can make sense of sounds? I will give a webinar on robot audition on Thursday, 3 Dec at 1pm UK Acoustics Network Plus Early Career SIG. Details below. Electronics Comp Sci MINDS CDT UKRI TAS Hub

New preprint: "Computational bioacoustics with deep learning: a review and roadmap" arxiv.org/abs/2112.06725 #bioacoustics #machinelearning

Hi guys / Hola amigos / Bom dia, Are you a researcher in a low-and-middle income country and working in #melanoma (any aspect of it)? Please get in touch, I would like to know more about what you're doing and your current challenges 😊 Please RT!!!

``ReCLAP: Improving Zero Shot Audio Classification by Describing Sounds,'' Sreyan Ghosh, Sonal Kumar, Chandra Kiran Reddy Evuru, Oriol Nieto, Ramani Duraiswami, Dinesh Manocha, ift.tt/OgxJruG

Important reminder 🚨: with the surge of "all 17 papers from my lab got accepted at NeurIPS 🤩" posts today, just remember that it's perfectly fine if your paper got rejected 🙅‍♀️ / if you haven't even submitted 🤷‍♀️. You can do awesome research even without flashy publications! 📝

You want to give audio abilities to your VLM without compromising its vision performance? You want to align your audio encoder with a pretrained image encoder without suffering from the modality gap? Check our #NeurIPS2024 paper with Michel Olvera Stéphane LATHUILIÈRE and Slim Essid

Great talk today by Haohe Liu at the ADASP group on Latent Diffusion Models (LDMs) as versatile audio decoder! Walked us through diffusion basics, AudioLDM for text-to-audio, audio quality enhancement, and neural codecs!

DCASE meets DCASE!? Michel Olvera

Thank you for coming to Tokyo, everyone! Let’s see next year at Barcelona!!!! #DCASE2024

Thank you for participating in #DCASE2024 workshop. Your participation made it a meaningful experience, and we look forward to staying connected.

🔊🎶 MIR folks, mirdata 0.3.9 is out‼️ ISMIR Conference #ISMIR2024 6 new loaders, for a total of 58 dataset loaders! This time, we made mirdata MUCH lighter! Install Time: 50.6s ⏩slashed by 31% to 34.6s Size: 95MB ⏬ shrunk by 98% to 1.9MB!! Datasets: mirdata.readthedocs.io/en/stable/sour…

🎉 New Pre-print! 🎉 Do CLIP models truly generalize to new, out-of-domain (OOD) images, or are they only doing well because they’ve been exposed to these domains in training? Our latest study reveals that CLIP’s ability to “generalize OOD” may be more limited than previously

If you want to learn more about audio-visual alignment and how to use it to give audio abilities to your VLM, stop by our NeurIPS Conference poster #3602 (East exhibit hall A-C) tomorrow at 11am!

An Eye for an Ear: Zero-shot Audio Description Leveraging an Image Captioner using Audiovisual Distribution Alignment