# Transfer Learning for Speech Recognition on a Budget

## TL;DR

This paper demonstrates that transfer learning enables efficient training of speech recognition models on limited resources, reducing data and computational costs while maintaining accuracy.

## Contribution

It introduces a transfer learning approach for adapting ASR models to new languages using less data and compute, especially suitable for resource-constrained environments.

## Key findings

- Transfer learning accelerates training on consumer hardware.
- Less training data needed for comparable accuracy.
- Small weight adaptations suffice for effective model transfer.

## Abstract

End-to-end training of automated speech recognition (ASR) systems requires massive data and compute resources. We explore transfer learning based on model adaptation as an approach for training ASR models under constrained GPU memory, throughput and training data. We conduct several systematic experiments adapting a Wav2Letter convolutional neural network originally trained for English ASR to the German language. We show that this technique allows faster training on consumer-grade resources while requiring less training data in order to achieve the same accuracy, thereby lowering the cost of training ASR models in other languages. Model introspection revealed that small adaptations to the network's weights were sufficient for good performance, especially for inner layers.