# Speech Model Pre-training for End-to-End Spoken Language Understanding

## TL;DR

This paper introduces a pre-training approach for end-to-end spoken language understanding that enhances performance and generalization, especially with limited training data, by predicting words and phonemes before fine-tuning on SLU tasks.

## Contribution

The paper proposes a novel pre-training method for end-to-end SLU models that reduces data requirements and improves accuracy, along with a new dataset for evaluation.

## Key findings

- Pre-training improves SLU accuracy with less data
- Method enhances generalization to unseen phrases
- New dataset facilitates SLU research

## Abstract

Whereas conventional spoken language understanding (SLU) systems map speech to text, and then text to intent, end-to-end SLU systems map speech directly to intent through a single trainable model. Achieving high accuracy with these end-to-end models without a large amount of training data is difficult. We propose a method to reduce the data requirements of end-to-end SLU in which the model is first pre-trained to predict words and phonemes, thus learning good features for SLU. We introduce a new SLU dataset, Fluent Speech Commands, and show that our method improves performance both when the full dataset is used for training and when only a small subset is used. We also describe preliminary experiments to gauge the model's ability to generalize to new phrases not heard during training.