Institution profile

Idiap Research Institute

Academic institutioneurope · ch
Official website
Research library216linked papers
Opportunities0open roles
Selected work

Representative Papers

Towards Open Diversity-Aware Social Interactions

Feb 17, 2025arXiv.org

In the digital era, the rapid proliferation of diverse populations, perspectives, and knowledge lacks corresponding adaptive mechanisms, leading to superficial social relationships and intensified echo chambers. Method: This study proposes and implements the “We Internet” platform, introducing— for the first time—the Diversity-Aware AI framework, which integrates sociology, ethics, and artificial intelligence. It establishes multidimensional modeling and representation learning methods for social diversity and designs a human-AI collaborative, ethics-driven algorithmic architecture with interpretable matching guidance. Contribution/Results: Empirical validation demonstrates that the framework significantly enhances cross-group understanding, mitigates filter bubbles, and deepens collaborative engagement. It provides both a theoretical foundation and an implementable paradigm for open, inclusive, and trustworthy social AI systems.

3 citationsRead paper

Text-only adaptation in LLM-based ASR through text denoising

Jan 28, 2026

This work addresses the performance degradation commonly observed in large language model (LLM)-based speech recognition systems when adapting solely with in-domain text, a process that often disrupts the alignment between speech and text modalities. To mitigate this issue, the authors propose a lightweight text-denoising adaptation approach that reformulates the audio projection task as a text denoising problem. By training the LLM to reconstruct clean transcripts from noisy textual inputs, the method achieves effective domain adaptation without modifying the model architecture or introducing additional parameters. This strategy preserves cross-modal alignment while significantly improving recognition accuracy, yielding up to a 22.1% relative reduction in word error rate on two benchmark datasets—substantially outperforming current state-of-the-art text-only adaptation techniques.

1 citationsRead paper

Reducing Prompt Sensitivity in LLM-based Speech Recognition Through Learnable Projection

Jan 28, 2026

This work addresses the instability and high sensitivity to prompt selection in existing large language model (LLM)-based speech recognition approaches that rely on fixed, manually crafted prompts. To overcome this limitation, the authors propose a model-agnostic, learnable prompt projection module that adaptively maps prompt embeddings into more effective regions of the LLM’s input space, without modifying the underlying LLM architecture. This approach significantly reduces prompt sensitivity and enhances recognition robustness and consistency. Experimental results across four benchmark datasets demonstrate that the proposed method not only consistently outperforms the best handcrafted prompts but also substantially mitigates performance variance, yielding more reliable and stable recognition outcomes.

1 citationsRead paper
Recent publications

Latest Papers