Training speech recognizers with unpaired speech and text – known as unsupervised speech recognition (UASR) – is a crucial step toward extending ASR to low-resource languages in the long-tail distribution and enabling multimodal learning from non-parallel data. However, existing approaches based on phones often rely on costly resources such as grapheme-to-phoneme converters (G2Ps) and struggle to generalize to languages with ambiguous phoneme boundaries due to training instability. In this paper, we address both challenges by introducing a syllable-level UASR framework based on masked language modeling, which avoids the need for G2P and the instability of GAN-based methods. Our approach achieves up to a 40% relative reduction in character error rate (CER) on LibriSpeech and generalizes effectively to low-resource languages that have remained particularly difficult for prior methods. Code will be released upon acceptance.

Large audio-language models (LALMs) show strong zero-shot performance on tasks such as audio question answering, but still lag behind specialized systems on discriminative tasks such as audio classification. Recent work shows that sparse subsets of attention heads can serve as discriminative feature extractors for few-shot classification, but existing methods weight all selected heads equally. We propose Class-Conditional Sparse Attention Vectors for Large Audio-Language Models, a few-shot classifier that learns class-conditional importance weights over attention heads. By treating heads as class-specific experts and aggregating them with reliability-weighted soft voting, CALM allows different classes to rely on different subsets of heads. Across diverse few-shot audio, audiovisual, visual, and spoofing benchmarks, CALM consistently outperforms state-of-the-art uniform-voting baselines, with gains of up to 14.52, 1.53, and 8.35 absolute points on audio, audiovisual, and spoofing tasks, respectively.

Existing supervised sign language recognition systems rely on an abundance of well-annotated data. Instead, an unsupervised speech-to-sign language recognition (SSR-U) system learns to translate between spoken and sign languages by observing only non-parallel speech and sign-language corpora. We propose speech2sign-U, a neural network-based approach capable of both character-level and word-level SSR-U. Our approach significantly outperforms baselines directly adapted from unsupervised speech recognition (ASR-U) models by as much as 50% recall@10 on several challenging American sign language corpora with various levels of sample sizes, vocabulary sizes, and audio and visual variability. The code is available at https://github.com/cactuswiththoughts/UnsupSpeech2Sign.gitcactuswiththoughts/UnsupSpeech2Sign.git.

Unsupervised speech recognition ({pasted macro ‘ASRU’}/) is the problem of learning automatic speech recognition (ASR) systems from unpaired speech-only and text-only corpora. While various algorithms exist to solve this problem, a theoretical framework is missing to study their properties and address such issues as sensitivity to hyperparameters and training instability. In this paper, we proposed a general theoretical framework to study the properties of {pasted macro ‘ASRU’}/ systems based on random matrix theory and the theory of neural tangent kernels. Such a framework allows us to prove various learnability conditions and sample complexity bounds of {pasted macro ‘ASRU’}/. Extensive {pasted macro ‘ASRU’}/ experiments on synthetic languages with three classes of transition graphs provide strong empirical evidence for our theory (code available at https://github.com/cactuswiththoughts/UnsupASRTheory.gitcactuswiththoughts/UnsupASRTheory.git).

Phonemes are defined by their relationship to words: changing a phoneme changes the word. Learning a phoneme inventory with little supervision has been a longstanding challenge with important applications to under-resourced speech technology. In this paper, we bridge the gap between the linguistic and statistical definition of phonemes and propose a novel neural discrete representation learning model for self-supervised learning of phoneme inventory with raw speech and word labels. Under mild assumptions, we prove that the phoneme inventory learned by our approach converges to the true one with an exponentially low error rate. Moreover, in experiments on TIMIT and Mboshi benchmarks, our approach consistently learns a better phoneme-level representation and achieves a lower error rate in a zero-resource phoneme recognition task than previous state-of-the-art self-supervised representation learning algorithms.

Event coreference resolution is critical to understand events in the growing number of online news with multiple modalities including text, video, speech, etc. However, the events and entities depicting in different modalities may not be perfectly aligned and can be difficult to annotate, which makes the task especially challenging with little supervision available. To address the above issues, we propose a supervised model based on attention mechanism and an unsupervised model based on statistical machine translation, capable of learning the relative importance of modalities for event coreference resolution. Experiments on a video multimedia event dataset show that our multimodal models outperform text-only systems in event coreference resolution tasks. A careful analysis reveals that the performance gain of the multimodal model especially under unsupervised settings comes from better learning of visually salient events.