Yiming Chen, Xianghu Yue, Xiaoxue Gao, Chen Zhang, Luis Fernando D’Haro, Robby T. Tan, Haizhou Li. Findings of the Association for Computational Linguistics: EMNLP 2024. 2024.

Self-supervised learning has achieved remarkable success for learning speech representations from unlabeled data. The masking strategy plays an important role in the self-supervised learning algorithm. Most of the masking techniques operate at a frame level. In linguistics, phone is the smallest unit of sound. Hence, we believe that a masking technique that operates at a phoneme level will effectively encode the phonotactic and prosodic constraints of a spoken language, thus eventually benefits the downstream speech recognition tasks. In this work, we explore a novel segmental masking strategy. Specifically, we mask phonetically motivated speech segments according to the phonetic segmentation in an utterance. By doing so, we implicitly incorporate the properties of a spoken language, such as phonotactic constraints and duration of phonetic segments, into the pre-training. Through extensive experiments, we confirm that the segmental masking strategy consistently outperforms the frame-based masking counterpart. We also further investigate the effect of segmental masking unit size, i.e. phoneme, phoneme span, and lexical word. This work presents an important finding about masking strategy in speech representation learning.

Text-to-speech (TTS) has been extensively studied for generating high-quality speech with textual inputs, playing a crucial role in various real-time applications. For real-world deployment, ensuring stable and timely generation in TTS models against minor input perturbations is of paramount importance. Therefore, evaluating the robustness of TTS models against such perturbations, commonly known as adversarial attacks, is highly desirable. In this paper, we propose TTSlow, a novel adversarial approach specifically tailored to slow down the speech generation process in TTS systems. To induce long TTS waiting time, we design novel efficiency-oriented adversarial loss to encourage endless generation process. TTSlow encompasses two attack strategies targeting both text inputs and speaker embedding. Specifically, we propose TTSlow-text, which utilizes a combination of homoglyphs-based and swap-based perturbations, along with TTSlow-spk, which employs a gradient optimization attack approach for speaker embedding. TTSlow serves as the first attack approach targeting a wide range of TTS models, including autoregressive and non-autoregressive TTS ones, thereby advancing exploration in audio security. Extensive experiments are conducted to evaluate the inference efficiency of TTS models, and in-depth analysis of generated speech intelligibility is performed using Gemini. The results demonstrate that TTSlow can effectively slow down two TTS models across three publicly available datasets. We are committed to releasing the source code upon acceptance, facilitating further research and benchmarking in this domain.

Class-incremental learning poses a significant challenge under an exemplar-free constraint, leading to catastrophic forgetting and sub-par incremental accuracy. Previous attempts have focused primarily on single-modality tasks, such as image classification or audio event classification. However, in the context of Audio-Visual Class-Incremental Learning (AVCIL), the effective integration and utilization of heterogeneous modalities, with their complementary and enhancing characteristics, remains largely unexplored. To bridge this gap, we propose the Multi-Modal Analytic Learning (MMAL) framework, an exemplar-free solution for AVCIL that employs a closed-form, linear approach. To be specific, MMAL introduces a modality fusion module that re-formulates the AVCIL problem through a Recursive Least-Square (RLS) perspective. Complementing this, a Modality-Specific Knowledge Compensation (MSKC) module is designed to further alleviate the under-fitting limitation intrinsic to analytic learning by harnessing individual knowledge from audio and visual modality in tandem. Comprehensive experimental comparisons with existing methods show that our proposed MMAL demonstrates superior performance with the accuracy of 76.71%, 78.98%, and 76.19% on AVE, Kinetics-Sounds, and VGGSounds100 datasets, respectively, setting new state-of-the-art AVCIL performance. Notably, compared to those memory-based methods, our MMAL, being an exemplar-free approach, provides good data privacy and can better leverage multi-modal information for improved incremental accuracy.

Sound Source Localization (SSL) enabling technology for applications such as surveillance and robotics. While traditional Signal Processing (SP)-based SSL methods provide analytic solutions under specific signal and noise assumptions, recent Deep Learning (DL)-based methods have significantly outperformed them. However, their success depends on extensive training data and substantial computational resources. Moreover, they often rely on large-scale annotated spatial data and may struggle to adapt to evolving sound classes. To mitigate these challenges, we propose a novel Class Incremental Learning (CIL) approach, termed SSL-CIL, which avoids serious accuracy degradation due to catastrophic forgetting by incrementally updating the DL-based SSL model through a closed-form analytic solution. In particular, data privacy is ensured since the learning process does not revisit any historical data (exemplar-free), which is more suitable for smart home scenarios. Empirical results in the public SSLR dataset demonstrate the superior performance of our proposal, achieving a localization accuracy of 90.9%, surpassing other competitive methods. Data and code: https://github.com/catherine-qian/SSL-Analytic-Class-Incremental-Learning-/tree/main

There has been a long-standing quest for a unified audio-visual-text model to enable various multimodal understanding tasks, which mimics the listening, seeing, and reading process of human beings. Humans tend to represent knowledge using two separate systems: one for representing verbal (textual) information and one for representing non-verbal (visual and auditory) information. These two systems can operate independently but can also interact with each other. Motivated by this understanding of human cognition, in this paper, we introduce CoAVT – a novel cognition-inspired Correlated Audio-Visual-Text pre-training model to connect the three modalities. It contains a joint audio-visual encoder that learns to encode audio-visual synchronization information together with the audio and visual content for non-verbal information, and a text encoder to handle textual input for verbal information. To bridge the gap between modalities, CoAVT employs a query encoder, which contains a set of learnable query embeddings and extracts the most informative audiovisual features of the corresponding text. Additionally, to leverage the correspondences between audio and vision with language respectively, we also establish the audio-text and visual-text bi-modal alignments upon the foundational audiovisual-text tri-modal alignment to enhance the multimodal representation learning. Finally, we jointly optimize CoAVT model with three multimodal objectives: contrastive loss, matching loss, and language modeling loss. Extensive experiments show that CoAVT can learn strong multimodal correlations and be generalized to various downstream tasks. CoAVT establishes new state-of-the-art performance on text-video retrieval tasks on AudioCaps for both zero-shot and fine-tuning settings, audio-visual event classification and audio-visual retrieval tasks on AudioSet and VGGSound. The results demonstrate the effectiveness and superiority of the proposed model for multimodal processing.