Investigation of Ensemble features of Self-Supervised Pretrained Models for Automatic Speech Recognition
This work addresses the problem of enhancing ASR accuracy for speech processing applications, but it is incremental as it combines existing SSL models rather than introducing a new method.
The paper tackled improving automatic speech recognition (ASR) by using an ensemble of self-supervised learning (SSL) models, specifically HuBERT, Wav2vec2.0, and WaveLM, to exploit complementary features, resulting in improved performance over individual models on the Librispeech(100h) and WSJ datasets.
Self-supervised learning (SSL) based models have been shown to generate powerful representations that can be used to improve the performance of downstream speech tasks. Several state-of-the-art SSL models are available, and each of these models optimizes a different loss which gives rise to the possibility of their features being complementary. This paper proposes using an ensemble of such SSL representations and models, which exploits the complementary nature of the features extracted by the various pretrained models. We hypothesize that this results in a richer feature representation and shows results for the ASR downstream task. To this end, we use three SSL models that have shown excellent results on ASR tasks, namely HuBERT, Wav2vec2.0, and WaveLM. We explore the ensemble of models fine-tuned for the ASR task and the ensemble of features using the embeddings obtained from the pre-trained models for a downstream ASR task. We get improved performance over individual models and pre-trained features using Librispeech(100h) and WSJ dataset for the downstream tasks.