# On Negative Sampling for Contrastive Audio-Text Retrieval

arXiv: https://arxiv.org/abs/2211.04070 · HTML: https://mlsanity.com/paper/2211.04070/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Nov 2022.

## Built on by (most cited first, 2 of 2)

- [DPC: Dual-Prompt Collaboration for Tuning Vision-Language Models](https://mlsanity.com/paper/2503.13443/) (2025)
- [Distance Sampling-based Paraphraser Leveraging ChatGPT for Text Data Manipulation](https://mlsanity.com/paper/2405.00367/) (2024)

## Abstract (arXiv)

This paper investigates negative sampling for contrastive learning in the context of audio-text retrieval. The strategy for negative sampling refers to selecting negatives (either audio clips or textual descriptions) from a pool of candidates for a positive audio-text pair. We explore sampling strategies via model-estimated within-modality and cross-modality relevance scores for audio and text samples. With a constant training setting on the retrieval system from [1], we study eight sampling strategies, including hard and semi-hard negative sampling. Experimental results show that retrieval performance varies dramatically among different strategies. Particularly, by selecting semi-hard negatives with cross-modality scores, the retrieval system gains improved performance in both text-to-audio and audio-to-text retrieval. Besides, we show that feature collapse occurs while sampling hard negatives with cross-modality scores.