Anti-Transfer Learning for Task Invariance in Convolutional Neural Networks for Speech Processing

TL;DR

This paper introduces anti-transfer learning for speech processing with CNNs, which discourages learning representations aligned with orthogonal tasks to improve target task invariance and accuracy.

Contribution

It proposes a novel anti-transfer learning method that penalizes similarity to orthogonal task representations, enhancing invariance and generalization in speech CNN models.

Findings

Anti-transfer improves invariance to orthogonal tasks.

It consistently enhances classification accuracy.

The method is applicable across multiple speech and audio tasks.

Abstract

We introduce the novel concept of anti-transfer learning for speech processing with convolutional neural networks. While transfer learning assumes that the learning process for a target task will benefit from re-using representations learned for another task, anti-transfer avoids the learning of representations that have been learned for an orthogonal task, i.e., one that is not relevant and potentially misleading for the target task, such as speaker identity for speech recognition or speech content for emotion recognition. In anti-transfer learning, we penalize similarity between activations of a network being trained and another one previously trained on an orthogonal task, which yields more suitable representations. This leads to better generalization and provides a degree of control over correlations that are spurious or undesirable, e.g. to avoid social bias. We have implemented anti-transfer for convolutional neural networks in different configurations with several similarity metrics and aggregation functions, which we evaluate and analyze with several speech and audio tasks and settings, using six datasets. We show that anti-transfer actually leads to the intended invariance to the orthogonal task and to more appropriate features for the target task at hand. Anti-transfer learning consistently improves classification accuracy in all test cases. While anti-transfer creates computation and memory cost at training time, there is relatively little computation cost when using pre-trained models for orthogonal tasks. Anti-transfer is widely applicable and particularly useful where a specific invariance is desirable or where trained models are available and labeled data for orthogonal tasks are difficult to obtain.