Moreno La Quatra
PhD · Assistant Professor · Deep Learning Researcher
I am an Assistant Professor (RTD-A) at Kore University of Enna, where I work on deep learning for speech processing and natural language understanding. My research looks at how neural networks can capture the details of human communication, from acoustic patterns in speech to semantic structures in text.
My work covers speech processing, natural language understanding, multimodal learning, and music analysis. I am particularly interested in building models that work well across languages, domains, and acoustic conditions. This has led me to explore self-supervised learning, cross-lingual transfer, and efficient neural architectures. When I'm not training models, I'm probably debugging why they won't converge.1
News
- 2027-05 Co-organizing DS4EIW 2027, the 6th International Workshop on Data Science for Equality, Inclusion and Well-being Challenges, at ICDE 2027. Copenhagen, Denmark
- 2026-12 Technical Program Co-chair and Co-Located Events Chair at SLT 2026 (IEEE Spoken Language Technology Workshop). Palermo, Italy
-
2026-10-24
Talk at Linux Day 2026 (in Italian): Liberi di usare l'AI, on open models, data, and running AI on your own computer.
- Slides
- Risorse e materiali
- Guida Linux/macOS to try local AI with Hermes and Ollama
-
2026-09-27
Five papers at INTERSPEECH 2026:
- Etiology-Aware Speech Language Models for Dysarthric Speech Recognition (oral)
- SSL-based Sequence Matching for Unsupervised Audio Retrieval
- Synthetic Pathological Speech at Scale: A Flow Matching Approach for Clinical Data Augmentation
- Hallucination Benchmark for Speech Foundation Models (long track)
- RT-SEMamba: Real-Time Speech Enhancement Mamba via Progressive Knowledge Distillation
- 2026-09-07 Demo paper at ECML PKDD 2026: A General-Purpose Platform for Reproducible Experimental Workflow Management and Multimodal Data Acquisition. Naples, Italy
-
2026-08-17
At ICPR 2026:
- Co-organizing the 2nd MCMI Workshop on Multi and Cross-Modal Information (August 21)
- Temporal Modeling with Feature Fusion for Autism Spectrum Disorder Detection from Skeletal Motion (main conference)
- Vision-Language Models for Gait-Based Screening via Zero-Shot and Multimodal In-Context Learning (workshops)
- 2026-05-11 Paper at LREC 2026: FAME: Fictional Actors for Multilingual Erasure. Palma, Mallorca, Spain
- 2026-05-04 Co-organizing the DS4EIW 2026 workshop on Data Science for Equality, Inclusion and Well-being at ICDE 2026. Montréal, Canada
- 2026-05-04 Paper at ICASSP 2026: When Audio Matters: A Lightweight, Hierarchical Fusion Model for Speech and Non-Verbal Emotion Recognition. Barcelona, Spain
- 2026-02-27 Paper accepted at EVALITA 2026: StereoBusters at GSI:detect on LLM-based detection and human analysis of gender stereotypes in Italian texts. Co-organizing the SVELA task on Machine Unlearning from LLM answers. Bari, Italy
- Archive 2025 · 2024 · 2023 · 2022
Master Thesis Supervision
If you are a student looking for a thesis topic in multimodal learning, audio processing, or natural language processing, feel free to contact me or browse the open thesis proposals for current topics.