Quan Wang

Shixuan Ma, Quan Wang

The increasing capability and widespread usage of large language models (LLMs) highlight the desirability of automatic detection of LLM-generated text. Zero-shot detectors, due to their training-free nature, have received considerable attention and notable success. In this paper, we identify a new feature, token cohesiveness, that is useful for zero-shot detection, and we demonstrate that LLM-generated text tends to exhibit higher token cohesiveness than human-written text. Based on this observation, we devise TOCSIN, a generic dual-channel detection paradigm that uses token cohesiveness as a plug-and-play module to improve existing zero-shot detectors. To calculate token cohesiveness, TOCSIN only requires a few rounds of random token deletion and semantic difference measurement, making it particularly suitable for a practical black-box setting where the source model used for generation is not accessible. Extensive experiments with four state-of-the-art base detectors on various datasets, source models, and evaluation settings demonstrate the effectiveness and generality of the proposed approach. Code available at: \url{https://github.com/Shixuan-Ma/TOCSIN}.

Beltrán Labrador, Pai Zhu, Guanlong Zhao et al.

Keyword spotting systems often struggle to generalize to a diverse population with various accents and age groups. To address this challenge, we propose a novel approach that integrates speaker information into keyword spotting using Feature-wise Linear Modulation (FiLM), a recent method for learning from multiple sources of information. We explore both Text-Dependent and Text-Independent speaker recognition systems to extract speaker information, and we experiment on extracting this information from both the input audio and pre-enrolled user audio. We evaluate our systems on a diverse dataset and achieve a substantial improvement in keyword detection accuracy, particularly among underrepresented speaker groups. Moreover, our proposed approach only requires a small 1% increase in the number of parameters, with a minimum impact on latency and computational cost, which makes it a practical solution for real-world applications.

Beltrán Labrador, Guanlong Zhao, Ignacio López Moreno et al.

In this paper, we present a novel approach to adapt a sequence-to-sequence Transformer-Transducer ASR system to the keyword spotting (KWS) task. We achieve this by replacing the keyword in the text transcription with a special token <kw> and training the system to detect the <kw> token in an audio stream. At inference time, we create a decision function inspired by conventional KWS approaches, to make our approach more suitable for the KWS task. Furthermore, we introduce a specific keyword spotting loss by adapting the sequence-discriminative Minimum Bayes-Risk training technique. We find that our approach significantly outperforms ASR based KWS systems. When compared with a conventional keyword spotting system, our proposal has similar performance while bringing the advantages and flexibility of sequence-to-sequence training. Additionally, when combined with the conventional KWS system, our approach can improve the performance at any operation point.

Yihan Chen, Benfeng Xu, Quan Wang et al.

While large language models (LLMs) have exhibited impressive instruction-following capabilities, it is still unclear whether and to what extent they can respond to explicit constraints that might be entailed in various instructions. As a significant aspect of LLM alignment, it is thus important to formulate such a specialized set of instructions as well as investigate the resulting behavior of LLMs. To address this vacancy, we propose a new benchmark CoDI-Eval to systematically and comprehensively evaluate LLMs' responses to instructions with various constraints. We construct a large collection of constraints-attributed instructions as a test suite focused on both generalization and coverage. Specifically, we advocate an instruction diversification process to synthesize diverse forms of constraint expression and also deliberate the candidate task taxonomy with even finer-grained sub-categories. Finally, we automate the entire evaluation process to facilitate further developments. Different from existing studies on controllable text generation, CoDI-Eval extends the scope to the prevalent instruction-following paradigm for the first time. We provide extensive evaluations of representative LLMs (e.g., ChatGPT, Vicuna) on CoDI-Eval, revealing their limitations in following instructions with specific constraints and there is still a significant gap between open-source and commercial closed-source LLMs. We believe this benchmark will facilitate research into improving the controllability of LLMs' responses to instructions. Our data and code are available at https://github.com/Xt-cyh/CoDI-Eval.

Jiahao Li, Quan Wang, Chiwei Zhu et al.

The task of Grammatical Error Correction (GEC) aims to automatically correct grammatical errors in natural texts. Almost all previous works treat annotated training data equally, but inherent discrepancies in data are neglected. In this paper, the inherent discrepancies are manifested in two aspects, namely, accuracy of data annotation and diversity of potential annotations. To this end, we propose MainGEC, which designs token-level and sentence-level training weights based on inherent discrepancies in accuracy and potential diversity of data annotation, respectively, and then conducts mixed-grained weighted training to improve the training effect for GEC. Empirical evaluation shows that whether in the Seq2Seq or Seq2Edit manner, MainGEC achieves consistent and significant performance improvements on two benchmark datasets, demonstrating the effectiveness and superiority of the mixed-grained weighted training. Further ablation experiments verify the effectiveness of designed weights of both granularities in MainGEC.

Shuai Liu, Di Wang, Quan Wang et al.

Previous single-stage detectors typically suffer the misalignment between localization accuracy and classification confidence. To solve the misalignment problem, we introduce a novel rectification method named neighbor IoU-voting (NIV) strategy. Typically, classification and regression are treated as separate branches, making it challenging to establish a connection between them. Consequently, the classification confidence cannot accurately reflect the regression quality. NIV strategy can serve as a bridge between classification and regression branches by calculating two types of statistical data from the regression output to correct the classification confidence. Furthermore, to alleviate the imbalance of detection accuracy for complete objects with dense points (easy objects) and incomplete objects with sparse points (difficult objects), we propose a new data augmentation scheme named object resampling. It undersamples easy objects and oversamples difficult objects by randomly transforming part of easy objects into difficult objects. Finally, combining the NIV strategy and object resampling augmentation, we design an efficient single-stage detector termed NIV-SSD. Extensive experiments on several datasets indicate the effectiveness of the NIV strategy and the competitive performance of the NIV-SSD detector. The code will be available at https://github.com/Say2L/NIV-SSD.

Chiwei Zhu, Benfeng Xu, An Yang et al.

Training language models with rationales augmentation has been shown to be beneficial in many existing works. In this paper, we identify that such a prevailing view does not hold consistently. We conduct comprehensive investigations to thoroughly inspect the impact of rationales on model performance as well as a novel perspective of model reliability. The results lead to several key findings that add new insights upon existing understandings: 1) Rationales can, at times, deteriorate model performance; 2) Rationales can, at times, improve model reliability, even outperforming their untrained counterparts; 3) A linear correspondence exists in between the performance and reliability improvements, while both are driven by the intrinsic difficulty of the task. These findings provide informative regulations on the broad utilization of rationales and raise critical implications on the procedure of explicitly aligning language models with implicit human thoughts. Codes can be found at https://github.com/Ignoramus0817/rationales.

Hui Li, Pengfei Yang, Juanyang Chen et al.

Knowledge distillation as an efficient knowledge transfer technique, has achieved remarkable success in unimodal scenarios. However, in cross-modal settings, conventional distillation methods encounter significant challenges due to data and statistical heterogeneities, failing to leverage the complementary prior knowledge embedded in cross-modal teacher models. This paper empirically reveals two critical issues in existing approaches: distillation path selection and knowledge drift. To address these limitations, we propose MST-Distill, a novel cross-modal knowledge distillation framework featuring a mixture of specialized teachers. Our approach employs a diverse ensemble of teacher models across both cross-modal and multimodal configurations, integrated with an instance-level routing network that facilitates adaptive and dynamic distillation. This architecture effectively transcends the constraints of traditional methods that rely on monotonous and static teacher models. Additionally, we introduce a plug-in masking module, independently trained to suppress modality-specific discrepancies and reconstruct teacher representations, thereby mitigating knowledge drift and enhancing transfer effectiveness. Extensive experiments across five diverse multimodal datasets, spanning visual, audio, and text, demonstrate that our method significantly outperforms existing state-of-the-art knowledge distillation methods in cross-modal distillation tasks. The source code is available at https://github.com/Gray-OREO/MST-Distill.

Mengqi Huang, Zhendong Mao, Quan Wang et al.

Existing autoregressive models follow the two-stage generation paradigm that first learns a codebook in the latent space for image reconstruction and then completes the image generation autoregressively based on the learned codebook. However, existing codebook learning simply models all local region information of images without distinguishing their different perceptual importance, which brings redundancy in the learned codebook that not only limits the next stage's autoregressive model's ability to model important structure but also results in high training cost and slow generation speed. In this study, we borrow the idea of importance perception from classical image coding theory and propose a novel two-stage framework, which consists of Masked Quantization VAE (MQ-VAE) and Stackformer, to relieve the model from modeling redundancy. Specifically, MQ-VAE incorporates an adaptive mask module for masking redundant region features before quantization and an adaptive de-mask module for recovering the original grid image feature map to faithfully reconstruct the original images after quantization. Then, Stackformer learns to predict the combination of the next code and its position in the feature map. Comprehensive experiments on various image generation validate our effectiveness and efficiency. Code will be released at https://github.com/CrossmodalGroup/MaskedVectorQuantization.

Quan Wang, Songtai Dai, Benfeng Xu et al.

Pre-trained language models (PLMs), such as BERT and GPT, have revolutionized the field of NLP, not only in the general domain but also in the biomedical domain. Most prior efforts in building biomedical PLMs have resorted simply to domain adaptation and focused mainly on English. In this work we introduce eHealth, a Chinese biomedical PLM built from scratch with a new pre-training framework. This new framework pre-trains eHealth as a discriminator through both token- and sequence-level discrimination. The former is to detect input tokens corrupted by a generator and recover their original identities from plausible candidates, while the latter is to further distinguish corruptions of a same original sequence from those of others. As such, eHealth can learn language semantics at both token and sequence levels. Extensive experiments on 11 Chinese biomedical language understanding tasks of various forms verify the effectiveness and superiority of our approach. We release the pre-trained model at \url{https://github.com/PaddlePaddle/Research/tree/master/KG/eHealth} and will also release the code later.

Beibin Li, Nicholas Nuechterlein, Erin Barney et al.

Identifying oculomotor behaviors relevant for eye-tracking applications is a critical but often challenging task. Aiming to automatically learn and extract knowledge from existing eye-tracking data, we develop a novel method that creates rich representations of oculomotor scanpaths to facilitate the learning of downstream tasks. The proposed stimulus-agnostic Oculomotor Behavior Framework (OBF) model learns human oculomotor behaviors from unsupervised and semi-supervised tasks, including reconstruction, predictive coding, fixation identification, and contrastive learning tasks. The resultant pre-trained OBF model can be used in a variety of applications. Our pre-trained model outperforms baseline approaches and traditional scanpath methods in autism spectrum disorder and viewed-stimulus classification tasks. Ablation experiments further show our proposed method could achieve even better results with larger model sizes and more diverse eye-tracking training datasets, supporting the model's potential for future eye-tracking applications. Open source code: http://github.com/BeibinLi/OBF.

Tianqi Zhong, Zhaoyi Li, Quan Wang et al.

Compositional generalization, representing the model's ability to generate text with new attribute combinations obtained by recombining single attributes from the training data, is a crucial property for multi-aspect controllable text generation (MCTG) methods. Nonetheless, a comprehensive compositional generalization evaluation benchmark of MCTG is still lacking. We propose CompMCTG, a benchmark encompassing diverse multi-aspect labeled datasets and a crafted three-dimensional evaluation protocol, to holistically evaluate the compositional generalization of MCTG approaches. We observe that existing MCTG works generally confront a noticeable performance drop in compositional testing. To mitigate this issue, we introduce Meta-MCTG, a training framework incorporating meta-learning, where we enable models to learn how to generalize by simulating compositional generalization scenarios in the training phase. We demonstrate the effectiveness of Meta-MCTG through achieving obvious improvement (by at most 3.64%) for compositional testing performance in 94.4% cases.

Guru Prakash Arumugam, Shuo-yiin Chang, Tara N. Sainath et al.

ASR models often suffer from a long-form deletion problem where the model predicts sequential blanks instead of words when transcribing a lengthy audio (in the order of minutes or hours). From the perspective of a user or downstream system consuming the ASR results, this behavior can be perceived as the model "being stuck", and potentially make the product hard to use. One of the culprits for long-form deletion is training-test data mismatch, which can happen even when the model is trained on diverse and large-scale data collected from multiple application domains. In this work, we introduce a novel technique to simultaneously model different groups of speakers in the audio along with the standard transcript tokens. Speakers are grouped as primary and non-primary, which connects the application domains and significantly alleviates the long-form deletion problem. This improved model neither needs any additional training data nor incurs additional training or inference cost.

Haodi Zhong, Liuxin Zou, Di Wang et al.

Dynamic graph-level embedding aims to capture structural evolution in networks, which is essential for modeling real-world scenarios. However, existing methods face two critical yet under-explored issues: Structural Visit Bias, where random walk sampling disproportionately emphasizes high-degree nodes, leading to redundant and noisy structural representations; and Abrupt Evolution Blindness, the failure to effectively detect sudden structural changes due to rigid or overly simplistic temporal modeling strategies, resulting in inconsistent temporal embeddings. To overcome these challenges, we propose EvoFormer, an evolution-aware Transformer framework tailored for dynamic graph-level representation learning. To mitigate Structural Visit Bias, EvoFormer introduces a Structure-Aware Transformer Module that incorporates positional encoding based on node structural roles, allowing the model to globally differentiate and accurately represent node structures. To overcome Abrupt Evolution Blindness, EvoFormer employs an Evolution-Sensitive Temporal Module, which explicitly models temporal evolution through a sequential three-step strategy: (I) Random Walk Timestamp Classification, generating initial timestamp-aware graph-level embeddings; (II) Graph-Level Temporal Segmentation, partitioning the graph stream into segments reflecting structurally coherent periods; and (III) Segment-Aware Temporal Self-Attention combined with an Edge Evolution Prediction task, enabling the model to precisely capture segment boundaries and perceive structural evolution trends, effectively adapting to rapid temporal shifts. Extensive evaluations on five benchmark datasets confirm that EvoFormer achieves state-of-the-art performance in graph similarity ranking, temporal anomaly detection, and temporal segmentation tasks, validating its effectiveness in correcting structural and temporal biases.

Xiao Liang, Di Wang, Zhicheng Jiao et al.

The rapid advancements in Vision Language Models (VLMs) have prompted the development of multi-modal medical assistant systems. Despite this progress, current models still have inherent probabilistic uncertainties, often producing erroneous or unverified responses-an issue with serious implications in medical applications. Existing methods aim to enhance the performance of Medical Vision Language Model (MedVLM) by adjusting model structure, fine-tuning with high-quality data, or through preference fine-tuning. However, these training-dependent strategies are costly and still lack sufficient alignment with clinical expertise. To address these issues, we propose an expert-in-the-loop framework named Expert-Controlled Classifier-Free Guidance (Expert-CFG) to align MedVLM with clinical expertise without additional training. This framework introduces an uncertainty estimation strategy to identify unreliable outputs. It then retrieves relevant references to assist experts in highlighting key terms and applies classifier-free guidance to refine the token embeddings of MedVLM, ensuring that the adjusted outputs are correct and align with expert highlights. Evaluations across three medical visual question answering benchmarks demonstrate that the proposed Expert-CFG, with 4.2B parameters and limited expert annotations, outperforms state-of-the-art models with 13B parameters. The results demonstrate the feasibility of deploying such a system in resource-limited settings for clinical use.

Xiao Liang, Jiawei Hu, Di Wang et al.

Vision-language models (VLMs) are prone to hallucinations that critically compromise reliability in medical applications. While preference optimization can mitigate these hallucinations through clinical feedback, its implementation faces challenges such as clinically irrelevant training samples, imbalanced data distributions, and prohibitive expert annotation costs. To address these challenges, we introduce CheXPO, a Chest X-ray Preference Optimization strategy that combines confidence-similarity joint mining with counterfactual rationale. Our approach begins by synthesizing a unified, fine-grained multi-task chest X-ray visual instruction dataset across different question types for supervised fine-tuning (SFT). We then identify hard examples through token-level confidence analysis of SFT failures and use similarity-based retrieval to expand hard examples for balancing preference sample distributions, while synthetic counterfactual rationales provide fine-grained clinical preferences, eliminating the need for additional expert input. Experiments show that CheXPO achieves 8.93% relative performance gain using only 5% of SFT samples, reaching state-of-the-art performance across diverse clinical tasks and providing a scalable, interpretable solution for real-world radiology applications.

Lei Han, Yitong Guo, Pengfei Yang et al.

Natural disasters have caused significant losses to human society, and the timely and efficient acquisition of post-disaster environmental information is crucial for the effective implementation of rescue operations. Due to the complexity of post-disaster environments, existing sensing technologies face challenges such as weak environmental adaptability, insufficient specialized sensing capabilities, and limited practicality of sensing solutions. This paper explores the heterogeneous multi-agent online autonomous collaborative scheduling algorithm HoAs-PALN, aimed at achieving efficient collection of post-disaster environmental information. HoAs-PALN is realized through adaptive dimensionality reduction in the matching process and local Nash equilibrium game, facilitating autonomous collaboration among time-dependent UAVs, workers and vehicles to enhance sensing scheduling. (1) In terms of adaptive dimensionality reduction during the matching process, HoAs-PALN significantly reduces scheduling decision time by transforming a five-dimensional matching process into two categories of three-dimensional matching processes; (2) Regarding the local Nash equilibrium game, HoAs-PALN combines the softmax function to optimize behavior selection probabilities and introduces a local Nash equilibrium determination mechanism to ensure scheduling decision performance. Finally, we conducted detailed experiments based on extensive real-world and simulated data. Compared with the baselines (GREEDY, K-WTA, MADL and MARL), HoAs-PALN improves task completion rates by 64.12%, 46.48%, 16.55%, and 14.03% on average, respectively, while each online scheduling decision takes less than 10 seconds, demonstrating its effectiveness in dynamic post-disaster environments.

Ye Jia, Michelle Tadmor Ramanovich, Quan Wang et al.

We introduce CVSS, a massively multilingual-to-English speech-to-speech translation (S2ST) corpus, covering sentence-level parallel S2ST pairs from 21 languages into English. CVSS is derived from the Common Voice speech corpus and the CoVoST 2 speech-to-text translation (ST) corpus, by synthesizing the translation text from CoVoST 2 into speech using state-of-the-art TTS systems. Two versions of translation speeches are provided: 1) CVSS-C: All the translation speeches are in a single high-quality canonical voice; 2) CVSS-T: The translation speeches are in voices transferred from the corresponding source speeches. In addition, CVSS provides normalized translation text which matches the pronunciation in the translation speech. On each version of CVSS, we built baseline multilingual direct S2ST models and cascade S2ST models, verifying the effectiveness of the corpus. To build strong cascade S2ST baselines, we trained an ST model on CoVoST 2, which outperforms the previous state-of-the-art trained on the corpus without extra data by 5.8 BLEU. Nevertheless, the performance of the direct S2ST models approaches the strong cascade baselines when trained from scratch, and with only 0.1 or 0.7 BLEU difference on ASR transcribed translation when initialized from matching ST models.

Rajeev Rikhye, Quan Wang, Qiao Liang et al.

In this paper, we introduce a streaming keyphrase detection system that can be easily customized to accurately detect any phrase composed of words from a large vocabulary. The system is implemented with an end-to-end trained automatic speech recognition (ASR) model and a text-independent speaker verification model. To address the challenge of detecting these keyphrases under various noisy conditions, a speaker separation model is added to the feature frontend of the speaker verification model, and an adaptive noise cancellation (ANC) algorithm is included to exploit cross-microphone noise coherence. Our experiments show that the text-independent speaker verification model largely reduces the false triggering rate of the keyphrase detection, while the speaker separation model and adaptive noise cancellation largely reduce false rejections.

Roza Chojnacka, Jason Pelecanos, Quan Wang et al.

In this paper, we describe SpeakerStew - a hybrid system to perform speaker verification on 46 languages. Two core ideas were explored in this system: (1) Pooling training data of different languages together for multilingual generalization and reducing development cycles; (2) A novel triage mechanism between text-dependent and text-independent models to reduce runtime cost and expected latency. To the best of our knowledge, this is the first study of speaker verification systems at the scale of 46 languages. The problem is framed from the perspective of using a smart speaker device with interactions consisting of a wake-up keyword (text-dependent) followed by a speech query (text-independent). Experimental evidence suggests that training on multiple languages can generalize to unseen varieties while maintaining performance on seen varieties. We also found that it can reduce computational requirements for training models by an order of magnitude. Furthermore, during model inference on English data, we observe that leveraging a triage framework can reduce the number of calls to the more computationally expensive text-independent system by 73% (and reduce latency by 59%) while maintaining an EER no worse than the text-independent setup.

Jason Pelecanos, Quan Wang, Ignacio Lopez Moreno

Many neural network speaker recognition systems model each speaker using a fixed-dimensional embedding vector. These embeddings are generally compared using either linear or 2nd-order scoring and, until recently, do not handle utterance-specific uncertainty. In this work we propose scoring these representations in a way that can capture uncertainty, enroll/test asymmetry and additional non-linear information. This is achieved by incorporating a 2nd-stage neural network (known as a decision network) as part of an end-to-end training regimen. In particular, we propose the concept of decision residual networks which involves the use of a compact decision network to leverage cosine scores and to model the residual signal that's needed. Additionally, we present a modification to the generalized end-to-end softmax loss function to target the separation of same/different speaker scores. We observed significant performance gains for the two techniques.

Benfeng Xu, Quan Wang, Yajuan Lyu et al.

Entities, as the essential elements in relation extraction tasks, exhibit certain structure. In this work, we formulate such structure as distinctive dependencies between mention pairs. We then propose SSAN, which incorporates these structural dependencies within the standard self-attention mechanism and throughout the overall encoding stage. Specifically, we design two alternative transformation modules inside each self-attention building block to produce attentive biases so as to adaptively regularize its attention flow. Our experiments demonstrate the usefulness of the proposed entity structure and the effectiveness of SSAN. It significantly outperforms competitive baselines, achieving new state-of-the-art results on three popular document-level relation extraction datasets. We further provide ablation and visualization to show how the entity structure guides the model for better relation extraction. Our code is publicly available.

Shaojin Ding, Quan Wang, Shuo-yiin Chang et al.

In this paper, we propose "personal VAD", a system to detect the voice activity of a target speaker at the frame level. This system is useful for gating the inputs to a streaming on-device speech recognition system, such that it only triggers for the target user, which helps reduce the computational cost and battery consumption, especially in scenarios where a keyword detector is unpreferable. We achieve this by training a VAD-alike neural network that is conditioned on the target speaker embedding or the speaker verification score. For each frame, personal VAD outputs the probabilities for three classes: non-speech, target speaker speech, and non-target speaker speech. Under our optimal setup, we are able to train a model with only 130K parameters that outperforms a baseline system where individually trained standard VAD and speaker recognition networks are combined to perform the same task.

Li Wan, Prashant Sridhar, Yang Yu et al.

In many scenarios of a language identification task, the user will specify a small set of languages which he/she can speak instead of a large set of all possible languages. We want to model such prior knowledge into the way we train our neural networks, by replacing the commonly used softmax loss function with a novel loss function named tuplemax loss. As a matter of fact, a typical language identification system launched in North America has about 95% users who could speak no more than two languages. Using the tuplemax loss, our system achieved a 2.33% error rate, which is a relative 39.4% improvement over the 3.85% error rate of standard softmax loss method.

Quan Wang, Hannah Muckenhirn, Kevin Wilson et al.

In this paper, we present a novel system that separates the voice of a target speaker from multi-speaker signals, by making use of a reference signal from the target speaker. We achieve this by training two separate neural networks: (1) A speaker recognition network that produces speaker-discriminative embeddings; (2) A spectrogram masking network that takes both noisy spectrogram and speaker embedding as input, and produces a mask. Our system significantly reduces the speech recognition WER on multi-speaker signals, with minimal WER degradation on single-speaker signals.

Aonan Zhang, Quan Wang, Zhenyao Zhu et al.

In this paper, we propose a fully supervised speaker diarization approach, named unbounded interleaved-state recurrent neural networks (UIS-RNN). Given extracted speaker-discriminative embeddings (a.k.a. d-vectors) from input utterances, each individual speaker is modeled by a parameter-sharing RNN, while the RNN states for different speakers interleave in the time domain. This RNN is naturally integrated with a distance-dependent Chinese restaurant process (ddCRP) to accommodate an unknown number of speakers. Our system is fully supervised and is able to learn from examples where time-stamped speaker labels are annotated. We achieved a 7.6% diarization error rate on NIST SRE 2000 CALLHOME, which is better than the state-of-the-art method using spectral clustering. Moreover, our method decodes in an online fashion while most state-of-the-art systems rely on offline clustering.

Yutian Chen, Yannis Assael, Brendan Shillingford et al.

We present a meta-learning approach for adaptive text-to-speech (TTS) with few data. During training, we learn a multi-speaker model using a shared conditional WaveNet core and independent learned embeddings for each speaker. The aim of training is not to produce a neural network with fixed weights, which is then deployed as a TTS system. Instead, the aim is to produce a network that requires few data at deployment time to rapidly adapt to new speakers. We introduce and benchmark three strategies: (i) learning the speaker embedding while keeping the WaveNet core fixed, (ii) fine-tuning the entire architecture with stochastic gradient descent, and (iii) predicting the speaker embedding with a trained neural network encoder. The experiments show that these approaches are successful at adapting the multi-speaker neural network to new speakers, obtaining state-of-the-art results in both sample naturalness and voice similarity with merely a few minutes of audio data from new speakers.

F A Rezaur Rahman Chowdhury, Quan Wang, Ignacio Lopez Moreno et al.

Attention-based models have recently shown great performance on a range of tasks, such as speech recognition, machine translation, and image captioning due to their ability to summarize relevant information that expands through the entire length of an input sequence. In this paper, we analyze the usage of attention mechanisms to the problem of sequence summarization in our end-to-end text-dependent speaker recognition system. We explore different topologies and their variants of the attention layer, and compare different pooling methods on the attention weights. Ultimately, we show that attention-based models can improves the Equal Error Rate (EER) of our speaker verification system by relatively 14% compared to our non-attention LSTM baseline model.

Quan Wang, Carlton Downey, Li Wan et al.

For many years, i-vector based audio embedding techniques were the dominant approach for speaker verification and speaker diarization applications. However, mirroring the rise of deep learning in various domains, neural network based audio embeddings, also known as d-vectors, have consistently demonstrated superior speaker verification performance. In this paper, we build on the success of d-vector based speaker verification systems to develop a new d-vector based approach to speaker diarization. Specifically, we combine LSTM-based d-vector audio embeddings with recent work in non-parametric clustering to obtain a state-of-the-art speaker diarization system. Our system is evaluated on three standard public datasets, suggesting that d-vector based diarization systems offer significant advantages over traditional i-vector based systems. We achieved a 12.0% diarization error rate on NIST SRE 2000 CALLHOME, while our model is trained with out-of-domain data from voice search logs.

Li Wan, Quan Wang, Alan Papir et al.

In this paper, we propose a new loss function called generalized end-to-end (GE2E) loss, which makes the training of speaker verification models more efficient than our previous tuple-based end-to-end (TE2E) loss function. Unlike TE2E, the GE2E loss function updates the network in a way that emphasizes examples that are difficult to verify at each step of the training process. Additionally, the GE2E loss does not require an initial stage of example selection. With these properties, our model with the new loss function decreases speaker verification EER by more than 10%, while reducing the training time by 60% at the same time. We also introduce the MultiReader technique, which allows us to do domain adaptation - training a more accurate model that supports multiple keywords (i.e. "OK Google" and "Hey Google") as well as multiple dialects.

Quan Wang, Kim L. Boyer

We present the MDS feature learning framework, in which multidimensional scaling (MDS) is applied on high-level pairwise image distances to learn fixed-length vector representations of images. The aspects of the images that are captured by the learned features, which we call MDS features, completely depend on what kind of image distance measurement is employed. With properly selected semantics-sensitive image distances, the MDS features provide rich semantic information about the images that is not captured by other feature extraction techniques. In our work, we introduce the iterated Levenberg-Marquardt algorithm for solving MDS, and study the MDS feature learning with IMage Euclidean Distance (IMED) and Spatial Pyramid Matching (SPM) distance. We present experiments on both synthetic data and real images --- the publicly accessible UIUC car image dataset. The MDS features based on SPM distance achieve exceptional performance for the car recognition task.