Singing voice beat and downbeat tracking posses several applications in automatic music production, analysis and manipulation. Among them, some require real-time processing, such as live performance processing and auto-accompaniment for singing inputs. This task is challenging owing to the non-trivial rhythmic and harmonic patterns in singing signals. For real-time processing, it introduces further constraints such as inaccessibility to future data and the impossibility to correct the previous results that are inconsistent with the latter ones. In this paper, we introduce the first system that tracks the beats and downbeats of singing voices in real-time. Specifically, we propose a novel dynamic particle filtering approach that incorporates offline historical data to correct the online inference by using a variable number of particles. We evaluate the performance on two datasets: GTZAN with the separated vocal tracks, and an in-house dataset with the original vocal stems. Experimental result demonstrates that our proposed approach outperforms the baseline by 3-5%.
Non-linguistic filler words, such as "uh" or "um", are prevalent in spontaneous speech and serve as indicators for expressing hesitation or uncertainty. Previous works for detecting certain non-linguistic filler words are highly dependent on transcriptions from a well-established commercial automatic speech recognition (ASR) system. However, certain ASR systems are not universally accessible from many aspects, e.g., budget, target languages, and computational power. In this work, we investigate filler word detection system that does not depend on ASR systems. We show that, by using the structured state space sequence model (S4) and neural semi-Markov conditional random fields (semi-CRFs), we achieve an absolute F1 improvement of 6.4% (segment level) and 3.1% (event level) on the PodcastFillers dataset. We also conduct a qualitative analysis on the detected results to analyze the limitations of our proposed system.
Voice anti-spoofing systems are crucial auxiliaries for automatic speaker verification (ASV) systems. A major challenge is caused by unseen attacks empowered by advanced speech synthesis technologies. Our previous research on one-class learning has improved the generalization ability to unseen attacks by compacting the bona fide speech in the embedding space. However, such compactness lacks consideration of the diversity of speakers. In this work, we propose speaker attractor multi-center one-class learning (SAMO), which clusters bona fide speech around a number of speaker attractors and pushes away spoofing attacks from all the attractors in a high-dimensional embedding space. For training, we propose an algorithm for the co-optimization of bona fide speech clustering and bona fide/spoof classification. For inference, we propose strategies to enable anti-spoofing for speakers without enrollment. Our proposed system outperforms existing state-of-the-art single systems with a relative improvement of 38% on equal error rate (EER) on the ASVspoof2019 LA evaluation set.
Head-related transfer functions (HRTFs) are a set of functions describing the spatial filtering effect of the outer ear (i.e., torso, head, and pinnae) onto sound sources at different azimuth and elevation angles. They are widely used in spatial audio rendering. While the azimuth and elevation angles are intrinsically continuous, measured HRTFs in existing datasets employ different spatial sampling schemes, making it difficult to model HRTFs across datasets. In this work, we propose to use neural fields, a differentiable representation of functions through neural networks, to model HRTFs with arbitrary spatial sampling schemes. Such representation is unified across datasets with different spatial sampling schemes. HRTFs for arbitrary azimuth and elevation angles can be derived from this representation. We further introduce a generative model named HRTF field to learn the latent space of the HRTF neural fields across subjects. We demonstrate promising performance on HRTF interpolation and generation tasks and point out potential future work.
Recent developments in neural speech synthesis and vocoding have sparked a renewed interest in voice conversion (VC). Beyond timbre transfer, achieving controllability on para-linguistic parameters such as pitch and rhythm is critical in deploying VC systems in many application scenarios. Existing studies, however, either only provide utterance-level global control or lack interpretability on the controls. In this paper, we propose ControlVC, the first neural voice conversion system that achieves time-varying controls on pitch and rhythm. ControlVC uses pre-trained encoders to compute pitch embeddings and linguistic embeddings from the source utterance and speaker embeddings from the target utterance. These embeddings are then concatenated and converted to speech using a vocoder. It achieves rhythm control through TD-PSOLA pre-processing on the source utterance, and achieves pitch control by manipulating the pitch contour before feeding it to the pitch encoder. Systematic subjective and objective evaluations are conducted to assess the speech quality and controllability. Results show that, on non-parallel and zero-shot conversion tasks, ControlVC significantly outperforms two other self-constructed baselines on speech quality, and it can successfully achieve time-varying pitch control.
Tracking beats of singing voices without the presence of musical accompaniment can find many applications in music production, automatic song arrangement, and social media interaction. Its main challenge is the lack of strong rhythmic and harmonic patterns that are important for music rhythmic analysis in general. Even for human listeners, this can be a challenging task. As a result, existing music beat tracking systems fail to deliver satisfactory performance on singing voices. In this paper, we propose singing beat tracking as a novel task, and propose the first approach to solving this task. Our approach leverages semantic information of singing voices by employing pre-trained self-supervised WavLM and DistilHuBERT speech representations as the front-end and uses a self-attention encoder layer to predict beats. To train and test the system, we obtain separated singing voices and their beat annotations using source separation and beat tracking on complete songs, followed by manual corrections. Experiments on the 741 separated vocal tracks of the GTZAN dataset show that the proposed system outperforms several state-of-the-art music beat tracking methods by a large margin in terms of beat tracking accuracy. Ablation studies also confirm the advantages of pre-trained self-supervised speech representations over generic spectral features.
In the growing field of virtual auditory display, personalized head-related transfer functions (HRTFs) play a vital role in establishing an accurate sound image. In this work, we propose an HRTF personalization method employing convolutional neural networks (CNN) to predict a subject's HRTFs for all directions from their scanned head geometry. To ease the training of the CNN models, we propose novel pre-processing methods for both the head scans and HRTF data to achieve compact representations. For the head scan, we use truncated spherical cap harmonic (SCH) coefficients to represent the pinna area, which is important in the acoustic scattering process. For the HRTF data, we use truncated spherical harmonic (SH) coefficients to represent the HRTF magnitudes and onsets. One CNN model is trained to predict the SH coefficients of the HRTF magnitudes from the SCH coefficients of the scanned ear geometry and other anthropometric measurements of the head. The other CNN model is trained to predict SH coefficients of the HRTF onsets from only the anthropometric measurements of the ear, head, and torso. Combining the magnitude and onset predictions, our method is able to predict the complete and global HRTF data. A leave-one-out validation with the log-spectral distortion (LSD) metric is used for objective evaluation. The results show a decent LSD level at both spatial \& temporal dimensions compared to the ground-truth HRTFs and a lower LSD than the boundary element method (BEM) simulation of HRTFs that the database provides. The localization simulation results with an auditory model are also consistent with the objective evaluation metrics, showing the localization responses with our predicted HRTFs are significantly better than with the BEM calculated ones.
Active speaker detection (ASD) systems are important modules for analyzing multi-talker conversations. They aim to detect which speakers or none are talking in a visual scene at any given time. Existing research on ASD does not agree on the definition of active speakers. We clarify the definition in this work and require synchronization between the audio and visual speaking activities. This clarification of definition is motivated by our extensive experiments, through which we discover that existing ASD methods fail in modeling the audio-visual synchronization and often classify unsynchronized videos as active speaking. To address this problem, we propose a cross-modal contrastive learning strategy and apply positional encoding in attention modules for supervised ASD models to leverage the synchronization cue. Experimental results suggest that our model can successfully detect unsynchronized speaking as not speaking, addressing the limitation of current models.
Full supervision models for source separation are trained on mixture-source parallel data and have achieved superior performance in recent years. However, large-scale and naturally mixed parallel training data are difficult to obtain for music, and such models are difficult to adapt to mixtures with new sources. Source-only supervision models, in contrast, only require clean sources for training; They learn source models and then apply these models to separate the mixture.