# Training Sound Event Detection On A Heterogeneous Dataset

arXiv: https://arxiv.org/abs/2007.03931 · HTML: https://mlsanity.com/paper/2007.03931/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Jul 2020.

## Built on by (most cited first, 7 of 7)

- [SEED: Sound Event Early Detection via Evidential Uncertainty](https://mlsanity.com/paper/2202.02441/) (2022)
- [Cross-Referencing Self-Training Network for Sound Event Detection in Audio Mixtures](https://mlsanity.com/paper/2105.13392/) (2021)
- [Synthetic data enables context-aware bioacoustic sound event detection](https://mlsanity.com/paper/2503.00296/) (2025)
- [Couple Learning for semi-supervised sound event detection](https://mlsanity.com/paper/2110.05809/) (2021)
- [A benchmark of state-of-the-art sound event detection systems evaluated on synthetic soundscapes](https://mlsanity.com/paper/2202.01487/) (2022)
- [Peer Collaborative Learning for Polyphonic Sound Event Detection](https://mlsanity.com/paper/2110.03511/) (2021)
- [Multi-encoder attention-based architectures for sound recognition with partial visual assistance](https://mlsanity.com/paper/2209.12826/) (2022)

## Abstract (arXiv)

Training a sound event detection algorithm on a heterogeneous dataset including both recorded and synthetic soundscapes that can have various labeling granularity is a non-trivial task that can lead to systems requiring several technical choices. These technical choices are often passed from one system to another without being questioned. We propose to perform a detailed analysis of DCASE 2020 task 4 sound event detection baseline with regards to several aspects such as the type of data used for training, the parameters of the mean-teacher or the transformations applied while generating the synthetic soundscapes. Some of the parameters that are usually used as default are shown to be sub-optimal.