Kexin Wang

Kexin Wang, Jiahong Zhang, Yong Ren et al.

Brain-inspired Spiking Neural Network (SNN) has demonstrated its effectiveness and efficiency in vision, natural language, and speech understanding tasks, indicating their capacity to "see", "listen", and "read". In this paper, we design \textbf{SpikeVoice}, which performs high-quality Text-To-Speech (TTS) via SNN, to explore the potential of SNN to "speak". A major obstacle to using SNN for such generative tasks lies in the demand for models to grasp long-term dependencies. The serial nature of spiking neurons, however, leads to the invisibility of information at future spiking time steps, limiting SNN models to capture sequence dependencies solely within the same time step. We term this phenomenon "partial-time dependency". To address this issue, we introduce Spiking Temporal-Sequential Attention STSA in the SpikeVoice. To the best of our knowledge, SpikeVoice is the first TTS work in the SNN field. We perform experiments using four well-established datasets that cover both Chinese and English languages, encompassing scenarios with both single-speaker and multi-speaker configurations. The results demonstrate that SpikeVoice can achieve results comparable to Artificial Neural Networks (ANN) with only 10.5 energy consumption of ANN.

Yuxi Heluo, Kexin Wang, Charles W. Robson

In this work, we contribute the first visual open-source empirical study on human behaviour during the COVID-19 pandemic, in order to investigate how compliant a general population is to mask-wearing-related public-health policy. Object-detection-based convolutional neural networks, regression analysis and multilayer perceptrons are combined to analyse visual data of the Viennese public during 2020. We find that mask-wearing-related government regulations and public-transport announcements encouraged correct mask-wearing-behaviours during the COVID-19 pandemic. Importantly, changes in announcement and regulation contents led to heterogeneous effects on people's behaviour. Comparing the predictive power of regression analysis and neural networks, we demonstrate that the latter produces more accurate predictions of population reactions during the COVID-19 pandemic. Our use of regression modelling also allows us to unearth possible causal pathways underlying societal behaviour. Since our findings highlight the importance of appropriate communication contents, our results will facilitate more effective non-pharmaceutical interventions to be developed in future. Adding to the literature, we demonstrate that regression modelling and neural networks are not mutually exclusive but instead complement each other.

Jiaan Wang, Jianfeng Qu, Kexin Wang et al.

Knowledge-grounded dialogue (KGD) learns to generate an informative response based on a given dialogue context and external knowledge (\emph{e.g.}, knowledge graphs; KGs). Recently, the emergence of large language models (LLMs) and pre-training techniques has brought great success to knowledge-grounded dialogue. However, when building KGD systems in real applications, there are various real-world noises that are inevitable to face. For example, the dialogue context might involve perturbations such as misspellings and abbreviations. In addition, KGs typically suffer from incompletion and also might contain erroneous and outdated facts. Such real-world noises pose a challenge to the robustness of KGD systems and hinder their applications in the real world. In this paper, we propose an entity-based contrastive learning framework for improving the robustness of KGD. Specifically, we make use of the entity information in a KGD sample to create both its positive and negative samples which involve semantic-irrelevant and semantic-relevant perturbations, respectively. The contrastive learning framework ensures the KGD model is aware of these two types of perturbations, thus generating informative responses with the potentially noisy inputs in real applications. Experimental results on three benchmark datasets show that our method achieves new state-of-the-art performance in terms of automatic evaluation scores, verifying its effectiveness and potentiality. Furthermore, we show that our method can generate better responses than comparison models in both the noisy and the few-shot settings.