相关论文: Remarks on Optimal Scores for Speaker Recognition

Minimum Bayes Risk (MBR) decoding has been shown to be a powerful alternative to beam search decoding in a variety of text generation tasks. MBR decoding selects a hypothesis from a pool of hypotheses that has the least expected risk under…

Modern speaker verification systems primarily rely on speaker embeddings, followed by verification based on cosine similarity between the embedding vectors of the enrollment and test utterances. While effective, these methods struggle with…

Minimum Bayes Risk (MBR) decoding optimizes output selection by maximizing the expected utility value of an underlying human distribution. While prior work has shown the effectiveness of MBR decoding through empirical evaluation, few…

Enhancing noisy speech is an important task to restore its quality and to improve its intelligibility. In traditional non-machine-learning (ML) based approaches the parameters required for noise reduction are estimated blindly from the…

This paper focuses on multi-enrollment speaker recognition which naturally occurs in the task of online speaker clustering, and studies the properties of different scoring back-ends in this scenario. First, we show that popular cosine…

Inferring the probability distribution of sentences or word sequences is a key process in natural language processing. While word-level language models (LMs) have been widely adopted for computing the joint probabilities of word sequences,…

Bayesian Neural Networks (BNNs) are trained to optimize an entire distribution over their weights instead of a single set, having significant advantages in terms of, e.g., interpretability, multi-task learning, and calibration. Because of…

We consider two questions at the heart of machine learning; how can we predict if a minimum will generalize to the test set, and why does stochastic gradient descent find minima that generalize well? Our work responds to Zhang et al.…

Minimum Bayes Risk (MBR) decoding is a text generation technique that has been shown to improve the quality of machine translations, but is expensive, even if a sampling-based approximation is used. Besides requiring a large number of…

Recently, an end-to-end speaker-attributed automatic speech recognition (E2E SA-ASR) model was proposed as a joint model of speaker counting, speech recognition and speaker identification for monaural overlapped speech. In the previous…

In nonlinear deterministic parameter estimation, the maximum likelihood estimator (MLE) is unable to attain the Cramer-Rao lower bound at low and medium signal-to-noise ratios (SNR) due the threshold and ambiguity phenomena. In order to…

Decision making is an important component in a speaker verification system. For the conventional GMM-UBM architecture, the decision is usually conducted based on the log likelihood ratio of the test utterance against the GMM of the claimed…

We introduce a new automatic evaluation method for speaker similarity assessment, that is consistent with human perceptual scores. Modern neural text-to-speech models require a vast amount of clean training data, which is why many solutions…

We introduce an information theoretic criterion for Bayesian network structure learning which we call quotient normalized maximum likelihood (qNML). In contrast to the closely related factorized normalized maximum likelihood criterion, qNML…

This paper investigates the effects of limited speech data in the context of speaker verification using deep neural network (DNN) approach. Being able to reduce the length of required speech data is important to the development of speaker…

Large language models (LLMs) have shown strong results on a range of applications, including regression and scoring tasks. Typically, one obtains outputs from an LLM via autoregressive sampling from the model's output distribution. We show…

When is optimal estimation linear? It is well known that, when a Gaussian source is contaminated with Gaussian noise, a linear estimator minimizes the mean square estimation error. This paper analyzes, more generally, the conditions for…

The prevailing noise-resistant and reverberation-resistant localization algorithms primarily emphasize separating and providing directional output for each speaker in multi-speaker scenarios, without association with the identity of…

In speaker recognition, where speech segments are mapped to embeddings on the unit hypersphere, two scoring backends are commonly used, namely cosine scoring or PLDA. Both have advantages and disadvantages, depending on the context. Cosine…

In this paper, we propose a simple but powerful unsupervised learning method for speaker recognition, namely Contrastive Equilibrium Learning (CEL), which increases the uncertainty on nuisance factors latent in the embeddings by employing…