# A Comparison of Methods for OOV-word Recognition on a New Public Dataset

arXiv: https://arxiv.org/abs/2107.08091 · HTML: https://mlsanity.com/paper/2107.08091/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 5 of its 24 references that are in the map.
- First posted on arXiv 16 Jul 2021.

## Builds on (most cited first, 5 of 5)

- [Common Voice: A Massively-Multilingual Speech Corpus](https://mlsanity.com/paper/1912.06670/) (2019)
- [ESPnet: End-to-End Speech Processing Toolkit](https://mlsanity.com/paper/1804.00015/) (2018)
- [wav2letter++: The Fastest Open-source Speech Recognition System](https://mlsanity.com/paper/1812.07625/) (2018)
- [Advanced Rich Transcription System for Estonian Speech](https://mlsanity.com/paper/1901.03601/) (2019)
- [Efficient Dynamic WFST Decoding for Personalized Language Models](https://mlsanity.com/paper/1910.10670/) (2019)

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

- [Improving callsign recognition with air-surveillance data in air-traffic communication](https://mlsanity.com/paper/2108.12156/) (2021)
- [Syllable Subword Tokens for Open Vocabulary Speech Recognition in Malayalam](https://mlsanity.com/paper/2301.06736/) (2023)

## Abstract (arXiv)

A common problem for automatic speech recognition systems is how to recognize words that they did not see during training. Currently there is no established method of evaluating different techniques for tackling this problem. We propose using the CommonVoice dataset to create test sets for multiple languages which have a high out-of-vocabulary (OOV) ratio relative to a training set and release a new tool for calculating relevant performance metrics. We then evaluate, within the context of a hybrid ASR system, how much better subword models are at recognizing OOVs, and how much benefit one can get from incorporating OOV-word information into an existing system by modifying WFSTs. Additionally, we propose a new method for modifying a subword-based language model so as to better recognize OOV-words. We showcase very large improvements in OOV-word recognition and make both the data and code available.