UniSLU: Unified Spoken Language Understanding from Heterogeneous Cross-Task Datasets
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866908454523764736 |
|---|---|
| author | Sheng, Zhichao Zhou, Shilin Gong, Chen Li, Zhenghua |
| author_facet | Sheng, Zhichao Zhou, Shilin Gong, Chen Li, Zhenghua |
| contents | Spoken Language Understanding (SLU) plays a crucial role in speech-centric multimedia applications, enabling machines to comprehend spoken language in scenarios such as meetings, interviews, and customer service interactions. SLU encompasses multiple tasks, including Automatic Speech Recognition (ASR), spoken Named Entity Recognition (NER), and spoken Sentiment Analysis (SA). However, existing methods often rely on separate model architectures for individual tasks such as spoken NER and SA, which increases system complexity, limits cross-task interaction, and fails to fully exploit heterogeneous datasets available across tasks. To address these limitations, we propose UniSLU, a unified framework that jointly models multiple SLU tasks within a single architecture. Specifically, we propose a unified representation for diverse SLU tasks, enabling full utilization of heterogeneous datasets across multiple tasks. Built upon this representation, we propose a unified generative method that jointly models ASR, spoken NER, and SA tasks, enhancing task interactions and enabling seamless integration with large language models to harness their powerful generative capabilities. Extensive experiments on public SLU datasets demonstrate the effectiveness of our approach, achieving superior SLU performance compared to several benchmark methods, making it well-suited for real-world speech-based multimedia scenarios. We will release all code and models at github to facilitate future research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_12951 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | UniSLU: Unified Spoken Language Understanding from Heterogeneous Cross-Task Datasets Sheng, Zhichao Zhou, Shilin Gong, Chen Li, Zhenghua Audio and Speech Processing Artificial Intelligence Computation and Language Multimedia Sound Spoken Language Understanding (SLU) plays a crucial role in speech-centric multimedia applications, enabling machines to comprehend spoken language in scenarios such as meetings, interviews, and customer service interactions. SLU encompasses multiple tasks, including Automatic Speech Recognition (ASR), spoken Named Entity Recognition (NER), and spoken Sentiment Analysis (SA). However, existing methods often rely on separate model architectures for individual tasks such as spoken NER and SA, which increases system complexity, limits cross-task interaction, and fails to fully exploit heterogeneous datasets available across tasks. To address these limitations, we propose UniSLU, a unified framework that jointly models multiple SLU tasks within a single architecture. Specifically, we propose a unified representation for diverse SLU tasks, enabling full utilization of heterogeneous datasets across multiple tasks. Built upon this representation, we propose a unified generative method that jointly models ASR, spoken NER, and SA tasks, enhancing task interactions and enabling seamless integration with large language models to harness their powerful generative capabilities. Extensive experiments on public SLU datasets demonstrate the effectiveness of our approach, achieving superior SLU performance compared to several benchmark methods, making it well-suited for real-world speech-based multimedia scenarios. We will release all code and models at github to facilitate future research. |
| title | UniSLU: Unified Spoken Language Understanding from Heterogeneous Cross-Task Datasets |
| topic | Audio and Speech Processing Artificial Intelligence Computation and Language Multimedia Sound |
| url | https://arxiv.org/abs/2507.12951 |