The European project PRESTO developed tools for broadcast archives to enable lower cost and more cost-effective digitisation of legacy and obsolete formats for audio and video. The required tools related to media selection, handling, conversion, long-term storage, distribution and access control.
Our pick of the week by @FBKZhihangXie: "#Speech Discrete Tokens or Continuous Features? A Comparative Analysis for Spoken Language Understanding in #SpeechLLMs" by @WangDingdo2603, Junan Li, @HelenMeng_CUHK, et al. (#EMNLP2025)
#SLU #SpeechTech
π New paper: Speech Discrete Tokens or Continuous Features?
π https://aclanthology.org/2025.emnlp-main.1266.pdf
π§© A comprehensive benchmark of SpeechLLMs using HuBERT/WavLM with Qwen & LLaMA.
β¨ Continuous features outperform overall, while discrete tokens excel at phoneme-level detail.
π Exciting news from the @FBK_MT group!
Four of our members @BeatriceSavoldi, @lina_conti, @negri_teo & @luisabentivogli are attending #EMNLP2025 in Suzhou π¨π³ with 5 accepted papers!
Come to our sessions & let's connect:
π https://mt.fbk.eu/fbk-mt-at-emnlp-2025/
Weβre also hiring postdocs!β‘
ππCongratulations to our PhD student @DennisFucci on a very successful thesis defense! π
Many thanks to the evaluation committee members @debora_nozza, @mirco_ravanelli, and Leonardo Badino for their insightful feedback and appreciation of his work!
#nlproc