Discrete tokens provide compact and domain-adaptable representations of speech features. However, their application to disordered speech, characterized by articulation imprecision and significant mismatch with normal voice, remains unexplored. To this end, this paper proposes novel phone-purity guided (PPG) discrete tokens to address the weakened phonetic discrimination arising during unsupervised K-means clustering or vector quantization of continuous features. Phonetic label supervision is incorporated to regularize the maximum likelihood and reconstruction error costs in standard K-means and VAE-VQ-based token extraction. Experiments on the UASpeech corpus show that PPG-based discrete tokens extracted from HuBERT consistently outperform hybrid TDNN and End-to-End (E2E) Conformer systems using non-PPG tokens. Statistically significant word error rate (WER) reductions of up to 0.99% and 1.77% absolute (3.21% and 4.82% relative) are achieved across varying codebook sizes for the 16 UASpeech test dysarthric speakers. The lowest WER of 23.25% is obtained by combining systems using complementary token features. Consistent improvements are also observed in phone purity, and t-SNE visualizations demonstrate sharper decision boundaries between K-means/VAE-VQ clusters with the introduction of phone-purity guidance.

This paper presents a novel mixed-precision quantization approach for speech foundation models that tightly integrates mixed-precision learning and quantized model parameter estimation into one single model compression stage. Experiments conducted on LibriSpeech dataset with fine-tuned wav2vec2.0-base and HuBERT-large models suggest the resulting mixed-precision quantized models increased the lossless compression ratio by factors up to 1.7x and 1.9x over the respective uniform-precision and two-stage mixed-precision quantized baselines that perform precision learning and model parameters quantization in separate and disjointed stages, while incurring no statistically word error rate (WER) increase over the 32-bit full-precision models. The system compression time of wav2vec2.0-base and HuBERT-large models is reduced by up to 1.9 and 1.5 times over the two-stage mixed-precision baselines, while both produce lower WERs. The best-performing 3.5-bit mixed-precision quantized HuBERT-large model produces a lossless compression ratio of 8.6x over the 32-bit full-precision system.

Interspeech 2024, 1-5 September 2024, Kos, Greece

In this paper, we consider the two-stage scheduling problem in which n jobs are first processed on m identical machines at a manufacturing facility and then delivered to their customers by one vehicle which can deliver one job at each shipment. In the problem, a set of n delivery times is given in advance, and in a schedule, the n delivery times should be assigned to the n jobs, respectively. The objective is to minimize the maximum delivery completion time, i.e., the time when all jobs are delivered to their respective customers and the vehicle returns to the facility. For this problem, we present a 32-approximation algorithm and a polynomial-time approximation scheme.

In this paper, we suggest a new optimization method by using grey area relational grade combined with nonlinear programming (abbreviated to NLP) method for GM(1,1) model’s parameters, we introduce the grey area relational grade to establish a NLP optimization parameters model for a GM(1,1) model that has been established, by using mathematical software LINGO 10.0 for its optimal solution. By a lot of data’s analysis, we conclude the new method is effective and feasible for the GM(1,1) models that have been established by using any other methods.