Xi Chen

Holly Wilson, Scott Wellington, Foteini Simistira Liwicki et al.

Decoding inner speech from the brain signal via hybridisation of fMRI and EEG data is explored to investigate the performance benefits over unimodal models. Two different bimodal fusion approaches are examined: concatenation of probability vectors output from unimodal fMRI and EEG machine learning models, and data fusion with feature engineering. Same task inner speech data are recorded from four participants, and different processing strategies are compared and contrasted to previously-employed hybridisation methods. Data across participants are discovered to encode different underlying structures, which results in varying decoding performances between subject-dependent fusion models. Decoding performance is demonstrated as improved when pursuing bimodal fMRI-EEG fusion strategies, if the data show underlying structure.

Ailin Huang, Bingxin Li, Bruce Wang et al.

Large Audio-Language Models (LALMs) have significantly advanced intelligent human-computer interaction, yet their reliance on text-based outputs limits their ability to generate natural speech responses directly, hindering seamless audio interactions. To address this, we introduce Step-Audio-AQAA, a fully end-to-end LALM designed for Audio Query-Audio Answer (AQAA) tasks. The model integrates a dual-codebook audio tokenizer for linguistic and semantic feature extraction, a 130-billion-parameter backbone LLM and a neural vocoder for high-fidelity speech synthesis. Our post-training approach employs interleaved token-output of text and audio to enhance semantic coherence and combines Direct Preference Optimization (DPO) with model merge to improve performance. Evaluations on the StepEval-Audio-360 benchmark demonstrate that Step-Audio-AQAA excels especially in speech control, outperforming the state-of-art LALMs in key areas. This work contributes a promising solution for end-to-end LALMs and highlights the critical role of token-based vocoder in enhancing overall performance for AQAA tasks.

Yifu Sun, Xulong Zhang, Yi Yu et al.

Singing voice detection (SVD), to recognize vocal parts in the song, is an essential task in music information retrieval (MIR). The task remains challenging since singing voice varies and intertwines with the accompaniment music, especially for some complicated polyphonic music such as choral music recordings. To address this problem, we investigate singing voice detection while discarding the interference from the accompaniment. The proposed SVD has two steps: i. The singing voice separation (SVS) technique is first utilized to filter out the singing voice's potential part coarsely. ii. Upon the continuity of vocal in the time domain, Long-term Recurrent Convolutional Networks (LRCN) is used to learn compositional features. Moreover, to eliminate the outliers, we choose to use a median filter for time-domain smoothing. Experimental results show that the proposed method outperforms the existing state-of-the-art works on two public datasets, the Jamendo Corpus and the RWC pop dataset.

Zhe Zhao, Hui Chen, Jinbin Zhang et al.

Existing works, including ELMO and BERT, have revealed the importance of pre-training for NLP tasks. While there does not exist a single pre-training model that works best in all cases, it is of necessity to develop a framework that is able to deploy various pre-training models efficiently. For this purpose, we propose an assemble-on-demand pre-training toolkit, namely Universal Encoder Representations (UER). UER is loosely coupled, and encapsulated with rich modules. By assembling modules on demand, users can either reproduce a state-of-the-art pre-training model or develop a pre-training model that remains unexplored. With UER, we have built a model zoo, which contains pre-trained models based on different corpora, encoders, and targets (objectives). With proper pre-trained models, we could achieve new state-of-the-art results on a range of downstream datasets.

Xi Chen, Hongzhi Wang, Yanjie Wei et al.

Time series prediction with missing values is an important problem of time series analysis since complete data is usually hard to obtain in many real-world applications. To model the generation of time series, autoregressive (AR) model is a basic and widely used one, which assumes that each observation in the time series is a noisy linear combination of some previous observations along with a constant shift. To tackle the problem of prediction with missing values, a number of methods were proposed based on various data models. For real application scenarios, how do these methods perform over different types of time series with different levels of data missing remains to be investigated. In this paper, we focus on online methods for AR-model-based time series prediction with missing values. We adapted five mainstream methods to fit in such a scenario. We make detailed discussion on each of them by introducing their core ideas about how to estimate the AR coefficients and their different strategies to deal with missing values. We also present algorithmic implementations for better understanding. In order to comprehensively evaluate these methods and do the comparison, we conduct experiments with various configurations of relative parameters over both synthetic and real data. From the experimental results, we derived several noteworthy conclusions and shows that imputation is a simple but reliable strategy to handle missing values in online prediction tasks.

Zongheng Yang, Eric Liang, Amog Kamsetty et al.

Cardinality estimation has long been grounded in statistical tools for density estimation. To capture the rich multivariate distributions of relational tables, we propose the use of a new type of high-capacity statistical model: deep autoregressive models. However, direct application of these models leads to a limited estimator that is prohibitively expensive to evaluate for range or wildcard predicates. To produce a truly usable estimator, we develop a Monte Carlo integration scheme on top of autoregressive models that can efficiently handle range queries with dozens of dimensions or more. Like classical synopses, our estimator summarizes the data without supervision. Unlike previous solutions, we approximate the joint data distribution without any independence assumptions. Evaluated on real-world datasets and compared against real systems and dominant families of techniques, our estimator achieves single-digit multiplicative error at tail, an up to 90$\times$ accuracy improvement over the second best method, and is space- and runtime-efficient.