Speaker verification (SV) provides billions of voice-enabled devices with access control, and ensures the security of voice-driven technologies. As a type of biometrics, it is necessary that SV is unbiased, with consistent and reliable performance across speakers irrespective of their demographic, social and economic attributes. Current SV evaluation practices are insufficient for evaluating bias: they are over-simplified and aggregate users, not representative of real-life usage scenarios, and consequences of errors are not accounted for. This paper proposes design guidelines for constructing SV evaluation datasets that address these short-comings. We propose a schema for grading the difficulty of utterance pairs, and present an algorithm for generating inclusive SV datasets. We empirically validate our proposed method in a set of experiments on the VoxCeleb1 dataset. Our results confirm that the count of utterance pairs/speaker, and the difficulty grading of utterance pairs have a significant effect on evaluation performance and variability. Our work contributes to the development of SV evaluation practices that are inclusive and fair.
相關內容
We revisit multiple hypothesis testing and propose a two-phase test, where each phase is a fixed-length test and the second-phase proceeds only if a reject option is decided in the first phase. We derive achievable error exponents of error probabilities under each hypothesis and show that our two-phase test bridges over fixed-length and sequential tests in the similar spirit of Lalitha and Javidi (ISIT, 2016) for binary hypothesis testing. Specifically, our test could achieve the performance close to a sequential test with the asymptotic complexity of a fixed-length test and such test is named the almost fixed-length test. Motivated by practical applications where the generating distribution under each hypothesis is \emph{unknown}, we generalize our results to the statistical classification framework of Gutman (TIT, 1989). We first consider binary classification and then generalize our results to $M$-ary classification. For both cases, we propose a two-phase test, derive achievable error exponents and demonstrate that our two-phase test bridges over fixed-length and sequential tests. In particular, for $M$-ary classification, no final reject option is required to achieve the same exponent as the sequential test of Haghifam, Tan, and Khisti (TIT, 2021). Our results generalize the design and analysis of the almost fixed-length test for binary hypothesis testing to broader and more practical families of $M$-ary hypothesis testing and statistical classification.
Single channel target speaker separation (TSS) aims at extracting a speaker's voice from a mixture of multiple talkers given an enrollment utterance of that speaker. A typical deep learning TSS framework consists of an upstream model that obtains enrollment speaker embeddings and a downstream model that performs the separation conditioned on the embeddings. In this paper, we look into several important but overlooked aspects of the enrollment embeddings, including the suitability of the widely used speaker identification embeddings, the introduction of the log-mel filterbank and self-supervised embeddings, and the embeddings' cross-dataset generalization capability. Our results show that the speaker identification embeddings could lose relevant information due to a sub-optimal metric, training objective, or common pre-processing. In contrast, both the filterbank and the self-supervised embeddings preserve the integrity of the speaker information, but the former consistently outperforms the latter in a cross-dataset evaluation. The competitive separation and generalization performance of the previously overlooked filterbank embedding is consistent across our study, which calls for future research on better upstream features.
Online education has gained an increasing importance over the last decade for providing affordable high-quality education to students worldwide. This has been further magnified during the global pandemic as more students switched to study online. The majority of online education tasks, e.g., course recommendation, exercise recommendation, or automated evaluation, depends on tracking students' knowledge progress. This is known as the \emph{Knowledge Tracing} problem in the literature. Addressing this problem requires collecting student evaluation data that can reflect their knowledge evolution over time. In this paper, we propose a new knowledge tracing dataset named Database Exercises for Knowledge Tracing (DBE-KT22) that is collected from an online student exercise system in a course taught at the Australian National University in Australia. We discuss the characteristics of the DBE-KT22 dataset and contrast it with the existing datasets in the knowledge tracing literature. Our dataset is available for public access through the Australian Data Archive platform.
In a sentence, certain words are critical for its semantic. Among them, named entities (NEs) are notoriously challenging for neural models. Despite their importance, their accurate handling has been neglected in speech-to-text (S2T) translation research, and recent work has shown that S2T models perform poorly for locations and notably person names, whose spelling is challenging unless known in advance. In this work, we explore how to leverage dictionaries of NEs known to likely appear in a given context to improve S2T model outputs. Our experiments show that we can reliably detect NEs likely present in an utterance starting from S2T encoder outputs. Indeed, we demonstrate that the current detection quality is sufficient to improve NE accuracy in the translation with a 31% reduction in person name errors.
Graph Neural Networks (GNNs) have been predominant for graph learning tasks; however, recent studies showed that a well-known graph algorithm, Label Propagation (LP), combined with a shallow neural network can achieve comparable performance to GNNs in semi-supervised node classification on graphs with high homophily. In this paper, we show that this approach falls short on graphs with low homophily, where nodes often connect to the nodes of the opposite classes. To overcome this, we carefully design a combination of a base predictor with LP algorithm that enjoys a closed-form solution as well as convergence guarantees. Our algorithm first learns the class compatibility matrix and then aggregates label predictions using LP algorithm weighted by class compatibilities. On a wide variety of benchmarks, we show that our approach achieves the leading performance on graphs with various levels of homophily. Meanwhile, it has orders of magnitude fewer parameters and requires less execution time. Empirical evaluations demonstrate that simple adaptations of LP can be competitive in semi-supervised node classification in both homophily and heterophily regimes.
Federated Learning (FL) is a distributed machine learning paradigm that enables learning models from decentralized private datasets, where the labeling effort is entrusted to the clients. While most existing FL approaches assume high-quality labels are readily available on users' devices; in reality, label noise can naturally occur in FL and follows a non-i.i.d. distribution among clients. Due to the non-iid-ness challenges, existing state-of-the-art centralized approaches exhibit unsatisfactory performance, while previous FL studies rely on data exchange or repeated server-side aid to improve model's performance. Here, we propose FedLN, a framework to deal with label noise across different FL training stages; namely, FL initialization, on-device model training, and server model aggregation. Specifically, FedLN computes per-client noise-level estimation in a single federated round and improves the models' performance by correcting (or limiting the effect of) noisy samples. Extensive experiments on various publicly available vision and audio datasets demonstrate a 24% improvement on average compared to other existing methods for a label noise level of 70%. We further validate the efficiency of FedLN in human-annotated real-world noisy datasets and report a 9% increase on average in models' recognition rate, highlighting that FedLN can be useful for improving FL services provided to everyday users.
The objective of this work is to develop a speaker recognition model to be used in diverse scenarios. We hypothesise that two components should be adequately configured to build such a model. First, adequate architecture would be required. We explore several recent state-of-the-art models, including ECAPA-TDNN and MFA-Conformer, as well as other baselines. Second, a massive amount of data would be required. We investigate several new training data configurations combining a few existing datasets. The most extensive configuration includes over 87k speakers' 10.22k hours of speech. Four evaluation protocols are adopted to measure how the trained model performs in diverse scenarios. Through experiments, we find that MFA-Conformer with the least inductive bias generalises the best. We also show that training with proposed large data configurations gives better performance. A boost in generalisation is observed, where the average performance on four evaluation protocols improves by more than 20%. In addition, we also demonstrate that these models' performances can improve even further when increasing capacity.
Knowledge enhanced pre-trained language models (K-PLMs) are shown to be effective for many public tasks in the literature but few of them have been successfully applied in practice. To address this problem, we propose K-AID, a systematic approach that includes a low-cost knowledge acquisition process for acquiring domain knowledge, an effective knowledge infusion module for improving model performance, and a knowledge distillation component for reducing the model size and deploying K-PLMs on resource-restricted devices (e.g., CPU) for real-world application. Importantly, instead of capturing entity knowledge like the majority of existing K-PLMs, our approach captures relational knowledge, which contributes to better-improving sentence-level text classification and text matching tasks that play a key role in question answering (QA). We conducted a set of experiments on five text classification tasks and three text matching tasks from three domains, namely E-commerce, Government, and Film&TV, and performed online A/B tests in E-commerce. Experimental results show that our approach is able to achieve substantial improvement on sentence-level question answering tasks and bring beneficial business value in industrial settings.
A community reveals the features and connections of its members that are different from those in other communities in a network. Detecting communities is of great significance in network analysis. Despite the classical spectral clustering and statistical inference methods, we notice a significant development of deep learning techniques for community detection in recent years with their advantages in handling high dimensional network data. Hence, a comprehensive overview of community detection's latest progress through deep learning is timely to both academics and practitioners. This survey devises and proposes a new taxonomy covering different categories of the state-of-the-art methods, including deep learning-based models upon deep neural networks, deep nonnegative matrix factorization and deep sparse filtering. The main category, i.e., deep neural networks, is further divided into convolutional networks, graph attention networks, generative adversarial networks and autoencoders. The survey also summarizes the popular benchmark data sets, model evaluation metrics, and open-source implementations to address experimentation settings. We then discuss the practical applications of community detection in various domains and point to implementation scenarios. Finally, we outline future directions by suggesting challenging topics in this fast-growing deep learning field.
User engagement is a critical metric for evaluating the quality of open-domain dialogue systems. Prior work has focused on conversation-level engagement by using heuristically constructed features such as the number of turns and the total time of the conversation. In this paper, we investigate the possibility and efficacy of estimating utterance-level engagement and define a novel metric, {\em predictive engagement}, for automatic evaluation of open-domain dialogue systems. Our experiments demonstrate that (1) human annotators have high agreement on assessing utterance-level engagement scores; (2) conversation-level engagement scores can be predicted from properly aggregated utterance-level engagement scores. Furthermore, we show that the utterance-level engagement scores can be learned from data. These scores can improve automatic evaluation metrics for open-domain dialogue systems, as shown by correlation with human judgements. This suggests that predictive engagement can be used as a real-time feedback for training better dialogue models.
小貼士
登錄享
相關主題
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191