# Estimated Audio-Caption Correspondences Improve Language-Based Audio Retrieval

arXiv: https://arxiv.org/abs/2408.11641 · HTML: https://mlsanity.com/paper/2408.11641/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 26 references that are in the map.
- First posted on arXiv 21 Aug 2024.

## Builds on (most cited first, 23 of 23)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [A Simple Framework for Contrastive Learning of Visual Representations](https://mlsanity.com/paper/2002.05709/) (2020)
- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [Pointer Sentinel Mixture Models](https://mlsanity.com/paper/1609.07843/) (2016)
- [Aligning Books and Movies: Towards Story-like Visual Explanations by Watching Movies and Reading Books](https://mlsanity.com/paper/1506.06724/) (2015)
- [Clotho: An Audio Captioning Dataset](https://mlsanity.com/paper/1910.09387/) (2019)
- [WavCaps: A ChatGPT-Assisted Weakly-Labelled Audio Captioning Dataset for Audio-Language Multimodal Research](https://mlsanity.com/paper/2303.17395/) (2023)
- [Efficient Training of Audio Transformers with Patchout](https://mlsanity.com/paper/2110.05069/) (2021)
- [VAST: A Vision-Audio-Subtitle-Text Omni-Modality Foundation Model and Dataset](https://mlsanity.com/paper/2305.18500/) (2023)
- [Audio Retrieval with Natural Language Queries: A Benchmark Study](https://mlsanity.com/paper/2112.09418/) (2021)
- [On Metric Learning for Audio-Text Cross-Modal Retrieval](https://mlsanity.com/paper/2203.15537/) (2022)
- [Efficient Large-scale Audio Tagging via Transformer-to-CNN Knowledge Distillation](https://mlsanity.com/paper/2211.04772/) (2022)
- [Self-supervised Audio Teacher-Student Transformer for Both Clip-level and Frame-level Tasks](https://mlsanity.com/paper/2306.04186/) (2023)
- [Cacophony: An Improved Contrastive Audio-Text Model](https://mlsanity.com/paper/2402.06986/) (2024)
- [Improving Text-Audio Retrieval by Text-aware Attention Pooling and Prior Matrix Revised Loss](https://mlsanity.com/paper/2303.05681/) (2023)
- [Unsupervised Audio-Caption Aligning Learns Correspondences between Individual Sound Events and Textual Phrases](https://mlsanity.com/paper/2110.02939/) (2021)
- [Advancing Natural-Language Based Audio Retrieval with PaSST and Large Audio-Caption Data Sets](https://mlsanity.com/paper/2308.04258/) (2023)
- [Killing two birds with one stone: Can an audio captioning system also be used for audio-text retrieval?](https://mlsanity.com/paper/2308.15090/) (2023)
- [Improving Natural-Language-based Audio Retrieval with Transfer Learning and Audio & Text Augmentations](https://mlsanity.com/paper/2208.11460/) (2022)
- [Fusing Audio and Metadata Embeddings Improves Language-based Audio Retrieval](https://mlsanity.com/paper/2406.15897/) (2024)
- [Crowdsourcing and Evaluating Text-Based Audio Retrieval Relevances](https://mlsanity.com/paper/2306.09820/) (2023)

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

- [TACOS: Temporally-aligned Audio CaptiOnS for Language-Audio Pretraining](https://mlsanity.com/paper/2505.07609/) (2025)
- [AISTAT lab system for DCASE2025 Task6: Language-based audio retrieval](https://mlsanity.com/paper/2509.16649/) (2025)

## Abstract (arXiv)

Dual-encoder-based audio retrieval systems are commonly optimized with contrastive learning on a set of matching and mismatching audio-caption pairs. This leads to a shared embedding space in which corresponding items from the two modalities end up close together. Since audio-caption datasets typically only contain matching pairs of recordings and descriptions, it has become common practice to create mismatching pairs by pairing the audio with a caption randomly drawn from the dataset. This is not ideal because the randomly sampled caption could, just by chance, partly or entirely describe the audio recording. However, correspondence information for all possible pairs is costly to annotate and thus typically unavailable; we, therefore, suggest substituting it with estimated correspondences. To this end, we propose a two-staged training procedure in which multiple retrieval models are first trained as usual, i.e., without estimated correspondences. In the second stage, the audio-caption correspondences predicted by these models then serve as prediction targets. We evaluate our method on the ClothoV2 and the AudioCaps benchmark and show that it improves retrieval performance, even in a restricting self-distillation setting where a single model generates and then learns from the estimated correspondences. We further show that our method outperforms the current state of the art by 1.6 pp. mAP@10 on the ClothoV2 benchmark.