This study focuses on emotion-sensitive spoken dialogue in human-machine speech interaction. With the advancement of Large Language Models (LLMs), dialogue systems can handle multimodal data, including audio. Recent models have enhanced the understanding of complex audio signals through the integration of various audio events. However, they are unable to generate appropriate responses based on emotional speech. To address this, we introduce the Emotional chat Model (E-chat), a novel spoken dialogue system capable of comprehending and responding to emotions conveyed from speech. This model leverages an emotion embedding extracted by a speech encoder, combined with LLMs, enabling it to respond according to different emotional contexts. Additionally, we introduce the E-chat200 dataset, designed explicitly for emotion-sensitive spoken dialogue. In various evaluation metrics, E-chat consistently outperforms baseline LLMs, demonstrating its potential in emotional comprehension and human-machine interaction.
Mixture-of-experts based models, which use language experts to extract language-specific representations effectively, have been well applied in code-switching automatic speech recognition. However, there is still substantial space to improve as similar pronunciation across languages may result in ineffective multi-language modeling and inaccurate language boundary estimation. To eliminate these drawbacks, we propose a cross-layer language adapter and a boundary-aware training method, namely Boundary-Aware Mixture-of-Experts (BA-MoE). Specifically, we introduce language-specific adapters to separate language-specific representations and a unified gating layer to fuse representations within each encoder layer. Second, we compute language adaptation loss of the mean output of each language-specific adapter to improve the adapter module's language-specific representation learning. Besides, we utilize a boundary-aware predictor to learn boundary representations for dealing with language boundary confusion. Our approach achieves significant performance improvement, reducing the mixture error rate by 16.55\% compared to the baseline on the ASRU 2019 Mandarin-English code-switching challenge dataset.
Multilingual automatic speech recognition (ASR) systems have garnered attention for their potential to extend language coverage globally. While self-supervised learning (SSL) has demonstrated its effectiveness in multilingual ASR, it is worth noting that the various layers' representations of SSL potentially contain distinct information that has not been fully leveraged. In this study, we propose a novel method that leverages self-supervised hierarchical representations (SSHR) to fine-tune multilingual ASR. We first analyze the different layers of the SSL model for language-related and content-related information, uncovering layers that show a stronger correlation. Then, we extract a language-related frame from correlated middle layers and guide specific content extraction through self-attention mechanisms. Additionally, we steer the model toward acquiring more content-related information in the final layers using our proposed Cross-CTC. We evaluate SSHR on two multilingual datasets, Common Voice and ML-SUPERB, and the experimental results demonstrate that our method achieves state-of-the-art performance to the best of our knowledge.
UniSpeech has achieved superior performance in cross-lingual automatic speech recognition (ASR) by explicitly aligning latent representations to phoneme units using multi-task self-supervised learning. While the learned representations transfer well from high-resource to low-resource languages, predicting words directly from these phonetic representations in downstream ASR is challenging. In this paper, we propose TranUSR, a two-stage model comprising a pre-trained UniData2vec and a phoneme-to-word Transcoder. Different from UniSpeech, UniData2vec replaces the quantized discrete representations with continuous and contextual representations from a teacher model for phonetically-aware pre-training. Then, Transcoder learns to translate phonemes to words with the aid of extra texts, enabling direct word generation. Experiments on Common Voice show that UniData2vec reduces PER by 5.3\% compared to UniSpeech, while Transcoder yields a 14.4\% WER reduction compared to grapheme fine-tuning.
With the development of the Generative Adversarial Networks (GANs) and DeepFakes, AI-synthesized images are now of such high quality that humans can hardly distinguish them from real images. It is imperative for media forensics to develop detectors to expose them accurately. Existing detection methods have shown high performance in generated images detection, but they tend to generalize poorly in the real-world scenarios, where the synthetic images are usually generated with unseen models using unknown source data. In this work, we emphasize the importance of combining information from the whole image and informative patches in improving the generalization ability of AI-synthesized image detection. Specifically, we design a two-branch model to combine global spatial information from the whole image and local informative features from multiple patches selected by a novel patch selection module. Multi-head attention mechanism is further utilized to fuse the global and local features. We collect a highly diverse dataset synthesized by 19 models with various objects and resolutions to evaluate our model. Experimental results demonstrate the high accuracy and good generalization ability of our method in detecting generated images.