Dennis Fucci

Speech Representation Analysis Based on Inter- and Intra-Model Similarities Self-supervised models have revolutionized speech processing, achieving new levels of performance in a wide variety of tasks with limited resources. However, the inner workings of these models are sti...

‪Findings from ieeexplore.ieee.org/document/106... show that speech SSL models converge on similar embedding spaces, but via different routes. While overall representations align, individual neurons learn distinct localized concepts.‬
‪Interesting read! @fbk-mt.bsky.social

🔍 Stiamo studiando come l'AI viene usata in Italia e per farlo abbiamo costruito un sondaggio!

👉 bit.ly/sondaggio_ai...

(è anonimo, richiede ~10 minuti, e se partecipi o lo fai girare ci aiuti un sacco🙏)

Ci interessa anche raggiungere persone che non si occupano e non sono esperte di AI!

Co-authored with:
@mgaido91.bsky.social, @matteo-negri.bsky.social, @mcettolo.bsky.social, @luisabentivogli.bsky.social

More great news! 🎉
Our paper “Echoes of Phonetics: Unveiling Relevant Acoustic Cues for ASR via Feature Attribution” has been accepted at #Interspeech2025!

If you’re interested in interpretability for speech models, stay tuned, the preprint is coming soon!

Big thanks to amazing co-authors: @mgaido91.bsky.social, @matteo-negri.bsky.social, @luisabentivogli.bsky.social, @andre-t-martins.bsky.social, @gattanasio.cc!

So happy our paper “Different Speech Translation Models Encode and Translate Speaker Gender Differently” was accepted at #ACL2025 (main)! 🎉

The preprint will be out soon!

#SpeechTranslation #GenderBias #Interpretability
@aclmeeting.bsky.social

📢 Come and join our group!
We offer a fully funded 3-year PhD position:

📔 Automatic translation with large multimodal models: iecs.unitn.it/education/ad...

📍Full details for application: iecs.unitn.it/education/ad...

📅 Deadline May 12, 2025

#NLProc #FBK

Had a great time visiting @milanlp.bsky.social last week and sharing some of my latest work on #XAI in #SpeechTranslation! Thanks for having me! 😊

Speech Is More than Words: Do Speech-to-Text Translation Systems Leverage Prosody? Ioannis Tsiamas, Matthias Sperber, Andrew Finch, Sarthak Garg. Proceedings of the Ninth Conference on Machine Translation. 2024.

As someone investigating how speech models use acoustic-related info, it’s great to see more work on evaluating whether prosodic features are leveraged in speech translation, like in aclanthology.org/2024.wmt-1.1.... We definitely need more research in this space!

@fbk-mt.bsky.social

🙌 All members of our group are now on Bluesky! 🙌

You can find all of us in this starter pack 👇