arXiv subjects

Muyong Cao

Publications and source records attributed to Muyong Cao.

4 recordsLinked to original sources

arXiv2025

LUPET: Incorporating Hierarchical Information Path into Multilingual ASR

Toward high-performance multilingual automatic speech recognition (ASR), various types of linguistic information and model design have demonstrated their effectiveness independently. They include language identity (LID), phoneme information, language-specific processing modules, and cross-lingual self-supervised speech representation. It is expected that leveraging their benefits synergistically in a unified solution would further improve the overall system performance. This paper presents a novel design of a hierarchical information path, named LUPET, which sequentially encodes, from the shallow layers to deep layers, multiple aspects of linguistic and acoustic information at diverse granularity scales. The path starts from LID prediction, followed by acoustic unit discovery, phoneme sharing, and finally token recognition routed by a mixture-of-expert. ASR experiments are carried out on 10 languages in the Common Voice corpus. The results demonstrate the superior performance of LUPET as compared to the baseline systems. Most importantly, LUPET effectively mitigates the issue of performance compromise of high-resource languages with low-resource ones in the multilingual setting.

eess.AS↗

arXiv2024

A Parameter-efficient Language Extension Framework for Multilingual ASR

Covering all languages with a multilingual speech recognition model (MASR) is very difficult. Performing language extension on top of an existing MASR is a desirable choice. In this study, the MASR continual learning problem is probabilistically decomposed into language identity prediction (LP) and cross-lingual adaptation (XLA) sub-problems. Based on this, we propose an architecture-based framework for language extension that can fundamentally solve catastrophic forgetting, debudded as PELE. PELE is designed to be parameter-efficient, incrementally incorporating an add-on module to adapt to a new language. Specifically, different parameter-efficient fine-tuning (PEFT) modules and their variants are explored as potential candidates to perform XLA. Experiments are carried out on 5 new languages with a wide range of low-resourced data sizes. The best-performing PEFT candidate can achieve satisfactory performance across all languages and demonstrates superiority in three of five languages over the continual joint learning setting. Notably, PEFT methods focusing on weight parameters or input features are revealed to be limited in performance, showing significantly inferior extension capabilities compared to inserting a lightweight module in between layers such as an Adapter.

cs.CL↗

arXiv2023

Fast and Blind Speech Copy-Move Detection and Localization in Noise

Copy-move forgery on speech (CMF), coupled with post-processing techniques, presents a great challenge to the forensic detection and localization of tampered areas. Most of the existing CMF detection approaches necessitate pre-segmentation of speech to facilitate similarity calculations among these segments. However, these approaches usually suffer from the problems of uncontrollable computational complexity and sensitivity to the presence of a word that is read multiple times within a speech recording. To address these issues, we propose a local feature tensors-based CMF detection algorithm that can transform duplicate detection and localization problems into a special tensor-matching procedure, accompanied by complete theoretical analysis as support. Through extensive experimentation, we have demonstrated that our method exhibits computational efficiency and robustness against post-processing techniques. Notably, it can effectively and blindly detect tampered segments, even those as short as a fractional second. These advantages highlight the promising potential of our approach for practical applications.

eess.AS↗

arXiv2022

Low-Complexity Acoustic Echo Cancellation with Neural Kalman Filtering

The Kalman filter has been adopted in acoustic echo cancellation due to its robustness to double-talk, fast convergence, and good steady-state performance. The performance of Kalman filter is closely related to the estimation accuracy of the state noise covariance and the observation noise covariance. The estimation error may lead to unacceptable results, especially when the echo path suffers abrupt changes, the tracking performance of the Kalman filter could be degraded significantly. In this paper, we propose the neural Kalman filtering (NKF), which uses neural networks to implicitly model the covariance of the state noise and observation noise and to output the Kalman gain in real-time. Experimental results on both synthetic test sets and real-recorded test sets show that, the proposed NKF has superior convergence and re-convergence performance while ensuring low near-end speech degradation comparing with the state-of-the-art model-based methods. Moreover, the model size of the proposed NKF is merely 5.3 K and the RTF is as low as 0.09, which indicates that it can be deployed in low-resource platforms.

eess.AS↗