Just ASR + LLM? A Study on Speech Large Language Models' Ability to Identify and Understand Speaker in Spoken Dialogue

In recent years, we have observed a rapid advancement in speech language models (SpeechLLMs), catching up with humans' listening and reasoning abilities. Remarkably, SpeechLLMs have demonstrated impressive spoken dialogue question-answering (SQA) performance in benchmarks like Gaokao, the English listening test of the college entrance exam in China, which seemingly requires understanding both the spoken content and voice characteristics of speakers in a conversation. However, after carefully examining Gaokao's questions, we find the correct answers to many questions can be inferred from the conversation context alone without identifying the speaker asked in the question. Our evaluation of state-of-the-art models Qwen-Audio and WavLLM in both Gaokao and our proposed "What Do You Like?" dataset shows a significantly higher accuracy in these context-based questions than in identity-critical questions, which can only be answered correctly with correct speaker identification. Our results and analysis suggest that when solving SQA, the current SpeechLLMs exhibit limited speaker awareness from the audio and behave similarly to an LLM reasoning from the conversation transcription without sound. We propose that our definitions and automated classification of context-based and identity-critical questions could offer a more accurate evaluation framework of SpeechLLMs in SQA tasks.

Meta-AF Echo Cancellation for Improved Keyword Spotting

Adaptive filters (AFs) are vital for enhancing the performance of downstream tasks, such as speech recognition, sound event detection, and keyword spotting. However, traditional AF design prioritizes isolated signal-level objectives, often overlooking downstream task performance. This can lead to suboptimal performance. Recent research has leveraged meta-learning to automatically learn AF update rules from data, alleviating the need for manual tuning when using simple signal-level objectives. This paper improves the Meta-AF framework by expanding it to support end-to-end training for arbitrary downstream tasks. We focus on classification tasks, where we introduce a novel training methodology that harnesses self-supervision and classifier feedback. We evaluate our approach on the combined task of acoustic echo cancellation and keyword spotting. Our findings demonstrate consistent performance improvements with both pre-trained and joint-trained keyword spotting models across synthetic and real playback. Notably, these improvements come without requiring additional tuning, increased inference-time complexity, or reliance on oracle signal-level training data.

Unsupervised Improvement of Audio-Text Cross-Modal Representations

Recent advances in using language models to obtain cross-modal audio-text representations have overcome the limitations of conventional training approaches that use predefined labels. This has allowed the community to make progress in tasks like zero-shot classification, which would otherwise not be possible. However, learning such representations requires a large amount of human-annotated audio-text pairs. In this paper, we study unsupervised approaches to improve the learning framework of such representations with unpaired text and audio. We explore domain-unspecific and domain-specific curation methods to create audio-text pairs that we use to further improve the model. We also show that when domain-specific curation is used in conjunction with a soft-labeled contrastive loss, we are able to obtain significant improvement in terms of zero-shot classification performance on downstream sound event classification or acoustic scene classification tasks.

Meta-Learning for Adaptive Filters with Higher-Order Frequency Dependencies

Adaptive filters are applicable to many signal processing tasks including acoustic echo cancellation, beamforming, and more. Adaptive filters are typically controlled using algorithms such as least-mean squares(LMS), recursive least squares(RLS), or Kalman filter updates. Such models are often applied in the frequency domain, assume frequency independent processing, and do not exploit higher-order frequency dependencies, for simplicity. Recent work on meta-adaptive filters, however, has shown that we can control filter adaptation using neural networks without manual derivation, motivating new work to exploit such information. In this work, we present higher-order meta-adaptive filters, a key improvement to meta-adaptive filters that incorporates higher-order frequency dependencies. We demonstrate our approach on acoustic echo cancellation and develop a family of filters that yield multi-dB improvements over competitive baselines, and are at least an order-of-magnitude less complex. Moreover, we show our improvements hold with or without a downstream speech enhancer.

Learning Representations for New Sound Classes With Continual Self-Supervised Learning

In this paper, we present a self-supervised learning framework for continually learning representations for new sound classes. The proposed system relies on a continually trained neural encoder that is trained with similarity-based learning objectives without using labels. We show that representations learned with the proposed method generalize better and are less susceptible to catastrophic forgetting than fully-supervised approaches. Remarkably, our technique does not store past data or models and is more computationally efficient than distillation-based methods. To accurately assess the system performance, in addition to using existing protocols, we propose two realistic evaluation protocols that use only a small amount of labeled data to simulate practical use cases.