Adapting Language Balance in Code-Switching Speech
This addresses the challenge of context bias in code-switching speech recognition, which is important for multilingual applications, but it is incremental as it builds on existing methods with a novel adaptation.
The paper tackled the problem of large foundational models struggling with code-switching speech due to infrequent code-switched moments, and by using a differentiable surrogate to highlight these points, it improved model robustness, reducing substitution errors in Arabic and Chinese-English experiments.
Despite achieving impressive results on standard benchmarks, large foundational models still struggle against code-switching test cases. When data scarcity cannot be used as the usual justification for poor performance, the reason may lie in the infrequent occurrence of code-switched moments, where the embedding of the second language appears subtly. Instead of expecting the models to learn this infrequency on their own, it might be beneficial to provide the training process with labels. Evaluating model performance on code-switching data requires careful localization of code-switching points where recognition errors are most consequential, so that the analysis emphasizes mistakes occurring at those moments. Building on this observation, we leverage the difference between the embedded and the main language to highlight those code-switching points and thereby emphasize learning at those locations. This simple yet effective differentiable surrogate mitigates context bias during generation -- the central challenge in code-switching -- thereby improving the model's robustness. Our experiments with Arabic and Chinese-English showed that the models are able to predict the switching places more correctly, reflected by the reduced substitution error.