Let's get a drink at the Stanislas place in the Nancy Opera House ! 🥂 The place just being a world UNESCO heritage site 🙄

🎹 pyannote + 🗒 notebook = pyannotebook pyannotebook is a custom Project Jupyter widget built on top of #pyannote.core and #wavesurferjs. It can be used to visualize and edit temporal audio labels, without leaving the notebook.

Etienne made a PyTorch package with datasets and dataloaders for Audio Captioning datasets (AudioCaps, Clotho, MACS for now): github.com/Labbeti/aac-da… pip install aac-datasets might be useful to AAC people

📣Following its successful evaluation, we anticipate a continuation of #ANITI in a form that we provisionally call 𝘈𝘕𝘐𝘛𝘐 2.0. ⚡️This call for chair will allow researchers to initiate the construction of their projects and submit their proposal ➡️aniti.univ-toulouse.fr/recherche-ia/c…

🫂Enjoying #pyannote speaker diarization pipelines? 👥What if it could do speaker separation as well? 🗣️Share the word if you want to make it happen! 🥳We are hiring! Twice! 👇

My language-based audio retrieval system, used in DCASE 2022 Task 6b, is now open-sourced: github.com/topel/IRIT-aud… Its originality was the use of AudioSet tags textual embeddings to describe the audio content Paper: ut3-toulouseinp.hal.science/hal-03812737/d…

We have 3 open 2-year PostDoc positions to work on ASR, sentiment analysis and driver's preferences modeling, for an in-car voice assistant, Come to work in Toulouse! ANITI Toulouse More info: irit.fr/~Thomas.Pelleg… #postdoc #postdocposition

Retraite : une génération de chercheuses et chercheurs très pénalisée. Les années de postdoc sont la plupart du temps des années blanches... Pour signer la pétition : chng.it/CxfpzC7g via Change.org France

Would you pay for #pyannote custom or premium models? #AskingForAFriend (I'd appreciate a few ReTweets 🙏)

I am considering starting an open-source business around #pyannote open-source state-of-the-art speaker diarization toolkit. Please help me make the right decisions by either filling this form and/or retweeting: forms.gle/eKhn7H2zTa68sM… A few promising stats:

🛠️ "New" audio tagging model on Hugging Face: hf.co/topel/ConvNeXt… - 28 M params - Trained on AudioSet - Test set mAP = 0.471 - provides 768-d global and frame-level embeddings Paper: isca-speech.org/archive/pdfs/i… #dcase #interspeech2023

Help needed! We are looking for reviewers for DCASE related topics for #ICASSP2024 ! If you have received an invitation, please respond to it. If you are not a reviewer but would like to be, contact Annamaria Mesaros or Romain Serizel to see about becoming one.

📢20 November 2023, 15h30, we're organising a keynote lecture and panel discussion with prof. Casilli. He will show how historical global inequalities shape international #digital labour and #data supply chains. 🖱️ Moderator Michele Mole 👉tinyurl.com/2p9bxtzc #AI

Not going to Seoul... Thanks to the reviewers who take care of my CO2 footprint! 🤷

📢 DCASE challenge 2024 short task descriptions are out! ⤵️ dcase.community/challenge2024/ Stay tuned for more details.

Machine listening people, please consider participating to the audio captioning task of DCASE. A new baseline system is provided: CNext-trans, 28M params, 29.6% SPIDEr-FL score on Clotho-eval dcase.community/challenge2024/… github.com/Labbeti/dcase2… #DCASE #audiocaptioning

#pyannote on iOS 📱/ macOS 💻 anyone? Call for tender has just been published (20 days left) achatpublic.com/sdm/ent2/gen/f… Please share if you'd like to see this happen 🙏

👀 Exciting advancements in Automated Audio Captioning.✨ The CMU-NVIDIA team has unveiled a groundbreaking approach that leverages multi-agent collaboration and GPU technology to enhance audio-to-text systems. 🔍 Key innovations include: ✅ Multi-encoder fusion: Combining

In this paper we show that is possible to create synthetic 2-speakers conversations with TTS and LLMs and fine-tune successfully Whisper for multi-speaker ASR generalizing well to real-world scenarios: arxiv.org/abs/2408.09215 Examples of such synth data: popcornell.github.io/SynthConvASRDe…