1 November 2018 in Computer Vision and Image Understanding
3 July 2024 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
Until recently, the field of speaker diarization was dominated by cascaded systems. Due to their limitations, mainly regarding overlapped speech and cumbersome pipelines, endto- end models have gained great popularity lately. One of the most successful models is end-to-end neural diarization with encoder-decoder based attractors (EEND-EDA). In this work, we replace the EDA module with a Perceiver-based one and show its advantages over EEND-EDA; namely obtaining better performance on the largely studied Callhome dataset, finding the quantity of speakers in a conversation more accurately, and faster inference time. Furthermore, when exhaustively compared with other methods, our model, DiaPer, reaches remarkable performance with a very lightweight design. Besides, we perform comparisons with other works and a cascaded baseline across more than ten public wide-band datasets. Together with this publication, we release the code of DiaPer as well as models trained on public and free data.
11 July 2019 in IEEE/ACM Transactions on Audio, Speech, and Language Processing
In this paper, we combine Hidden Markov Models (HMMs) with i-vector extractors to address the problem of text-dependent speaker recognition with random digit strings. We employ digit-specific HMMs to segment the utterances into digits, to perform frame alignment to HMM states and to extract Baum-Welch statistics. By making use of the natural partition of input features into digits, we train digit-specific i-vector extractors on top of each HMM and we extract well-localized i-vectors, each modelling merely the phonetic content corresponding to a single digit. We then examine ways to perform channel and uncertainty compensation, and we propose a novel method for using the uncertainty in the i-vector estimates. The experiments on RSR2015 part III show that the proposed method attains 1.52% and 1.77% Equal Error Rate (EER) for male and female respectively, outperforming state-of-the-art methods such as x-vectors, trained on vast amounts of data. Furthermore, these results are attained by a single system trained entirely on RSR2015, and by a simple score-normalized cosine distance. Moreover, we show that the omission of channel compensation yields only a minor degradation in performance, meaning that the system attains state-of-the-art results even without recordings from multiple handsets per speaker for training or enrolment. Similar conclusions are drawn from our experiments on the RedDots corpus, where the same method is evaluated on phrases. Finally, we report results with bottleneck features and show that further improvement is attained when fusing them with spectral features.
19 April 2010 in IEEE Journal of Selected Topics in Signal Processing
This paper discusses the use of the BIC with respect to speaker diarization, i.e., the problem of assigning the observation vectors of an audio file to a set of speakers of unknown cardinality. Our primary goals are to examine the two dominant approaches of the BIC, namely the global and the local and combine the strengths of the two variants into one intuitive criterion, the segmental-BIC. We then consider the asymptotic behavior of the segmental-BIC, when dealing with models that are highly misspecified, as the ones commonly used in the speaker diarization task. Our main result is a modified version of the BIC, which significantly outperforms the current variants over the entire range of operating points, and achieves performance close to those of highly computationally demanding algorithms.