Christos Plachouras 🦋

researcher & musician, phd student @c4dm (audio-language models), follow me on

[Christos Plachouras 🦋 @plachouras]

We launched a way to search for sounds on the internet with your voice 🎤✨ - just imitate what you're thinking of! It's free and open-source! Check it out: https://t.co/FbykxDGd2m. It uses a tiny AI model we made that runs on your browser, not a server 😉

[R. Oğuz Araz @r_oguz_araz]

Excited to be at #ISMIR2025 🇰🇷 this week presenting our recent work on Neural Music Fingerprinting!

🎶 SOTA results on real & synthetic datasets
🎶 Open-sourced code, pretrained models & data

📄 Preprint: https://t.co/dzyIVEmmCY
💻 Code: https://t.co/bzGTCJC2cw

[Christos Plachouras 🦋 @plachouras]

If you are at #IJCNN in Rome, see you tomorrow (Thu July 3) at 4:30pm in the Position Papers session, Leonardo da Vinci hall!

[Christos Plachouras 🦋 @plachouras]

Excited to share our new paper: https://t.co/kKK0OAKQqf! We introduce a unified framework for evaluating model representations beyond downstream tasks, and use it to uncover some interesting insights about the structure of representations that challenge conventional wisdom 🔍🧵

[Christos Plachouras 🦋 @plachouras]

More in the paper and code: https://t.co/K0VDVvAd0n. Big thanks to my collaborators @Juj_Guinot, George Fazekas, @elio_elioo, @emmanouilb, and Johan Pauwels! I’ll be at IJCNN 2025 in Rome in a month to present this - see you there!

[Christos Plachouras 🦋 @plachouras]

Excited to share our new paper: https://t.co/kKK0OAKQqf! We introduce a unified framework for evaluating model representations beyond downstream tasks, and use it to uncover some interesting insights about the structure of representations that challenge conventional wisdom 🔍🧵

[Christos Plachouras 🦋 @plachouras]

We argue that downstream task evaluation cannot easily uncover these behaviors, and that equivariance, invariance, and disentanglement are critical components that enable a variety of real-world applications like retrieval, generation, and style transfer.

[Christos Plachouras 🦋 @plachouras]

more at the audio coding and modeling poster sessions (AASP in poster area 2B) tomorrow (Friday) 8:30am to 10am! paper link: https://t.co/7LOHENMmZB

[Christos Plachouras 🦋 @plachouras]

how much music data do we actually need to pretrain effective music representation learning models? we decided to systematically investigate this in our latest #ICASSP2025 paper with @emmanouilb and Johan Pauwels

[Christos Plachouras 🦋 @plachouras]

finally, we found that regardless of pretraining data scale, the inherent robustness of representations to relevant data perturbations is still concerning

(MusiCNN 🟠, VGG 🔵, AST 🟢, CLMR 🔴, TMAE 🟣, MFCC 🩶)

[Marius Miron @nkundiushuti]

I am at ICASSP and will be presenting biodenoising on Friday morning.happy to talk to people interested in bioacoustics,cross-domain representation transfer or simply curious about our work at ESP.we released a new version of biodenoising including self-training on your own data

[Christos Plachouras 🦋 @plachouras]

🎶 5 years ago...

[Christos Plachouras 🦋 @plachouras]

very cool work significantly improving robust audio identification with graph neural nets: https://t.co/nvTBo5ofx9