Clotho

The 'Clotho' dataset is a benchmark that contains audio clips paired with descriptive text, used to evaluate the performance of multi-modal audio-language models in understanding and relating audio content to natural language.

Papers using Clotho (7)

FORTE: FOL-guided Optimal Refinement for Text-audio rEtrieval2026OmniRetriever: Any-to-Any Audio-Video-Text Retrieval via Fusion-as-Teacher Distillation2026AC/DC: LLM-based Audio Comprehension via Dialogue Continuation2025ONE-PEACE: Exploring One General Representation Model Toward Unlimited Modalities2023 · 42 citesZero-Shot Audio Captioning Using Soft and Hard Prompts2024 · 1 citesMINT: a Multi-modal Image and Narrative Text Dubbing Dataset for Foley Audio Content Planning and Generation2024Enhancing Audio-Language Models through Self-Supervised Post-Training with Text-Audio Pairs2024