Papers by Paraskevas Stamatiadis
iKnow-audio: Integrating Knowledge Graphs with Audio-Language Models (2025.emnlp-main)
Copied to clipboard
| Challenge: | Contrastive language-audio pretraining models learn by aligning audio and text in a shared embedding space. |
| Approach: | They propose a framework that integrates knowledge graphs with audio-language models to provide robust semantic grounding. |
| Outcome: | iKnow-audio improves disambiguation of acoustically similar sounds and reduces prompt engineering. |