Related papers: Efficiently Train ASR Models that Memorize Less an…

Full-parameter fine-tuning of large language models is constrained by substantial GPU memory requirements. Low-rank adaptation methods mitigate this challenge by updating only a subset of parameters. However, these approaches often limit…

Continual learning for automatic speech recognition (ASR) systems poses a challenge, especially with the need to avoid catastrophic forgetting while maintaining performance on previously learned tasks. This paper introduces a novel approach…

Each round in Differential Private Stochastic Gradient Descent (DPSGD) transmits a sum of clipped gradients obfuscated with Gaussian noise to a central server which uses this to update a global model which often represents a deep neural…

The recent emergence of joint CTC-Attention model shows significant improvement in automatic speech recognition (ASR). The improvement largely lies in the modeling of linguistic information by decoder. The decoder joint-optimized with an…

The neural transducer is an end-to-end model for automatic speech recognition (ASR). While the model is well-suited for streaming ASR, the training process remains challenging. During training, the memory requirements may quickly exceed the…

Predictive coding (PC) is a general theory of cortical function. The local, gradient-based learning rules found in one kind of PC model have recently been shown to closely approximate backpropagation. This finding suggests that this…

Discrete speech representations have garnered recent attention for their efficacy in training transformer-based models for various speech-related tasks such as automatic speech recognition (ASR), translation, speaker verification, and joint…

Recent advances have substantially improved the accuracy, memory cost, and training speed of differentially private (DP) deep learning, especially on large vision and language models with millions to billions of parameters. In this work, we…

Automatic Speech Recognition (ASR) has increased in popularity in recent years. The evolution of processor and storage technologies has enabled more advanced ASR mechanisms, fueling the development of virtual assistants such as Amazon…

Building a good speech recognition system usually requires large amounts of transcribed data, which is expensive to collect. To tackle this problem, many unsupervised pre-training methods have been proposed. Among these methods, Masked…

Loss spikes remain a persistent obstacle in large-scale language model pretraining. While previous research has attempted to identify the root cause of loss spikes by investigating individual factors, we observe that, in practice, such…

The precise programming of crossbar arrays of unit-cells is crucial for obtaining high matrix-vector-multiplication (MVM) accuracy in analog in-memory computing (AIMC) cores. We propose a radically different approach based on directly…

Previous research has shown that constraining the gradient of loss function with respect to model-predicted probabilities can enhance the model robustness against noisy labels. These methods typically specify a fixed optimal threshold for…

This paper studies a novel pre-training technique with unpaired speech data, Speech2C, for encoder-decoder based automatic speech recognition (ASR). Within a multi-task learning framework, we introduce two pre-training tasks for the…

The performance of automatic speech recognition systems under noisy environments still leaves room for improvement. Speech enhancement or feature enhancement techniques for increasing noise robustness of these systems usually add components…

Parameterized Quantum Circuits (PQC) are drawing increasing research interest thanks to its potential to achieve quantum advantages on near-term Noisy Intermediate Scale Quantum (NISQ) hardware. In order to achieve scalable PQC learning,…

Recent advancement in deep learning encouraged developing large automatic speech recognition (ASR) models that achieve promising results while ignoring computational and memory constraints. However, deploying such models on low resource…

Achieving communication efficiency in decentralized machine learning has been attracting significant attention, with communication compression recognized as an effective technique in algorithm design. This paper takes a first step to…

Automatic speech recognition (ASR) systems often encounter difficulties in accurately recognizing rare words, leading to errors that can have a negative impact on downstream tasks such as keyword spotting, intent detection, and text…

Modern ASR systems are typically trained on large-scale pseudo-labeled, in-the-wild data spanning multiple domains. While such heterogeneous data benefit generalist models designed for broad deployment, they pose challenges for specialist…