Search

Consistency regularization is a commonly used practice to encourage the model to generate consistent representation from distorted input features and improve model generalization. It shows significant improvement on various speech applications that are optimized with cross entropy criterion. However, it is not straightforward to apply consistency regularization for the transducer-based approaches, which are widely adopted for speech applications due to the competitive performance and streaming characteristic. The main challenge is from the vast alignment space of the transducer optimization criterion and not all the alignments within the space contribute to the model optimization equally. In this study, we present Transducer Consistency Regularization (TCR), a consistency regularization method for transducer models. We apply distortions such as spec augmentation and dropout to create different data views and minimize the distribution difference. We utilize occupational probabilities to give different weights on transducer output distributions, thus only alignments close to oracle alignments would contribute to the model learning. Our experiments show the proposed method is superior to other consistency regularization implementations and could effectively reduce word error rate (WER) by 4.3\% relatively comparing with a strong baseline on the \textsc{Librispeech} dataset., Comment: 8 pages, 4 figures. Accepted in IEEE Spoken Language Technology Workshop 2024

6G Open Radio Access Networks (O-RAN) promises to open data interfaces to enable plug-and-play service Apps, many of which are consumer and business-facing. Opening up 6G access lowers the barrier to innovation but raises the challenge that the required communication specifications are not fully known to all service designers. As such, business innovators must either be familiar with 6G standards or consult with experts. Enabling consistent, unbiased, rapid, and low-cost requirement assessment and specification generation is crucial to the O-RAN innovation ecosystem. Here, we discuss our initiative to bridge service specification gaps between network service providers and business innovators leveraging Large Language Models (LLMs). We first review the state-of-the-art and motivation in 6G plug-and-play services, capabilities, potential use cases and LLMs. We identify an ample innovation space for hybrid use cases that may require diverse and variational wireless functionalities across its operating time. We show that the network specification can be automated and present the first automatic retrieval-augmented network service specification framework for 6G use cases. To enable public acceptance and feedback, a website interface is published for the research and industrial community to experiment with the framework. We hope this review highlights the need for emerging foundation models for this area and motivates researcher engagement and contribution to the community through our framework., Comment: 7 pages, 6 figures, Submitted to IEEE Comm Magazine for possible publication

Text-to-Image (T2I) Diffusion Models (DMs) have shown impressive abilities in generating high-quality images based on simple text descriptions. However, as is common with many Deep Learning (DL) models, DMs are subject to a lack of robustness. While there are attempts to evaluate the robustness of T2I DMs as a binary or worst-case problem, they cannot answer how robust in general the model is whenever an adversarial example (AE) can be found. In this study, we first introduce a probabilistic notion of T2I DMs' robustness; and then establish an efficient framework, ProTIP, to evaluate it with statistical guarantees. The main challenges stem from: i) the high computational cost of the generation process; and ii) determining if a perturbed input is an AE involves comparing two output distributions, which is fundamentally harder compared to other DL tasks like classification where an AE is identified upon misprediction of labels. To tackle the challenges, we employ sequential analysis with efficacy and futility early stopping rules in the statistical testing for identifying AEs, and adaptive concentration inequalities to dynamically determine the "just-right" number of stochastic perturbations whenever the verification target is met. Empirical experiments validate the effectiveness and efficiency of ProTIP over common T2I DMs. Finally, we demonstrate an application of ProTIP to rank commonly used defence methods., Comment: Accepted by ECCV24

Engineering knowledge-based (or expert) systems require extensive manual effort and domain knowledge. As Large Language Models (LLMs) are trained using an enormous amount of cross-domain knowledge, it becomes possible to automate such engineering processes. This paper presents an empirical automation and semi-automation framework for domain knowledge distillation using prompt engineering and the LLM ChatGPT. We assess the framework empirically in the autonomous driving domain and present our key observations. In our implementation, we construct the domain knowledge ontology by "chatting" with ChatGPT. The key finding is that while fully automated domain ontology construction is possible, human supervision and early intervention typically improve efficiency and output quality as they lessen the effects of response randomness and the butterfly effect. We, therefore, also develop a web-based distillation assistant enabling supervision and flexible intervention at runtime. We hope our findings and tools could inspire future research toward revolutionizing the engineering of knowledge-based systems across application domains., Comment: Accepted by ITSC 2023

Hidden-unit BERT (HuBERT) is a widely-used self-supervised learning (SSL) model in speech processing. However, we argue that its fixed 20ms resolution for hidden representations would not be optimal for various speech-processing tasks since their attributes (e.g., speaker characteristics and semantics) are based on different time scales. To address this limitation, we propose utilizing HuBERT representations at multiple resolutions for downstream tasks. We explore two approaches, namely the parallel and hierarchical approaches, for integrating HuBERT features with different resolutions. Through experiments, we demonstrate that HuBERT with multiple resolutions outperforms the original model. This highlights the potential of utilizing multiple resolutions in SSL models like HuBERT to capture diverse information from speech signals., Comment: Accepted to Interspeech2023

Transducer and Attention based Encoder-Decoder (AED) are two widely used frameworks for speech-to-text tasks. They are designed for different purposes and each has its own benefits and drawbacks for speech-to-text tasks. In order to leverage strengths of both modeling methods, we propose a solution by combining Transducer and Attention based Encoder-Decoder (TAED) for speech-to-text tasks. The new method leverages AED's strength in non-monotonic sequence to sequence learning while retaining Transducer's streaming property. In the proposed framework, Transducer and AED share the same speech encoder. The predictor in Transducer is replaced by the decoder in the AED model, and the outputs of the decoder are conditioned on the speech inputs instead of outputs from an unconditioned language model. The proposed solution ensures that the model is optimized by covering all possible read/write scenarios and creates a matched environment for streaming applications. We evaluate the proposed approach on the \textsc{MuST-C} dataset and the findings demonstrate that TAED performs significantly better than Transducer for offline automatic speech recognition (ASR) and speech-to-text translation (ST) tasks. In the streaming case, TAED outperforms Transducer in the ASR task and one ST direction while comparable results are achieved in another translation direction., Comment: ACL 2023 main conference

It has been known that direct speech-to-speech translation (S2ST) models usually suffer from the data scarcity issue because of the limited existing parallel materials for both source and target speech. Therefore to train a direct S2ST system, previous works usually utilize text-to-speech (TTS) systems to generate samples in the target language by augmenting the data from speech-to-text translation (S2TT). However, there is a limited investigation into how the synthesized target speech would affect the S2ST models. In this work, we analyze the effect of changing synthesized target speech for direct S2ST models. We find that simply combining the target speech from different TTS systems can potentially improve the S2ST performances. Following that, we also propose a multi-task framework that jointly optimizes the S2ST system with multiple targets from different TTS systems. Extensive experiments demonstrate that our proposed framework achieves consistent improvements (2.8 BLEU) over the baselines on the Fisher Spanish-English dataset.

ESPnet-ST-v2 is a revamp of the open-source ESPnet-ST toolkit necessitated by the broadening interests of the spoken language translation community. ESPnet-ST-v2 supports 1) offline speech-to-text translation (ST), 2) simultaneous speech-to-text translation (SST), and 3) offline speech-to-speech translation (S2ST) -- each task is supported with a wide variety of approaches, differentiating ESPnet-ST-v2 from other open source spoken language translation toolkits. This toolkit offers state-of-the-art architectures such as transducers, hybrid CTC/attention, multi-decoders with searchable intermediates, time-synchronous blockwise CTC/attention, Translatotron models, and direct discrete unit models. In this paper, we describe the overall design, example models for each task, and performance benchmarking behind ESPnet-ST-v2, which is publicly available at https://github.com/espnet/espnet., Comment: ACL 2023; System Demonstration

It has become a consensus that autonomous vehicles (AVs) will first be widely deployed on highways. However, the complexity of highway interchanges becomes the bottleneck for deploying AVs. An AV should be sufficiently tested under different highway interchanges, which is still challenging due to the lack of available datasets containing diverse highway interchanges. In this paper, we propose a model-driven method, FLYOVER, to generate a dataset consisting of diverse interchanges with measurable diversity coverage. First, FLYOVER proposes a labeled digraph to model the topology of an interchange. Second, FLYOVER takes real-world interchanges as input to guarantee topology practicality and extracts different topology equivalence classes by classifying the corresponding topology models. Third, for each topology class, FLYOVER identifies the corresponding geometrical features for the ramps and generates concrete interchanges using k-way combinatorial coverage and differential evolution. To illustrate the diversity and applicability of the generated interchange dataset, we test the built-in traffic flow control algorithm in SUMO and the fuel-optimization trajectory tracking algorithm deployed to Alibaba's autonomous trucks on the dataset. The results show that except for the geometrical difference, the interchanges are diverse in throughput and fuel consumption under the traffic flow control and trajectory tracking algorithms, respectively., Comment: Accepted by ICRA 2023

Direct speech-to-speech translation (S2ST), in which all components can be optimized jointly, is advantageous over cascaded approaches to achieve fast inference with a simplified pipeline. We present a novel two-pass direct S2ST architecture, UnitY, which first generates textual representations and predicts discrete acoustic units subsequently. We enhance the model performance by subword prediction in the first-pass decoder, advanced two-pass decoder architecture design and search strategy, and better training regularization. To leverage large amounts of unlabeled text data, we pre-train the first-pass text decoder based on the self-supervised denoising auto-encoding task. Experimental evaluations on benchmark datasets at various data scales demonstrate that UnitY outperforms a single-pass speech-to-unit translation model by 2.5-4.2 ASR-BLEU with 2.83x decoding speed-up. We show that the proposed methods boost the performance even when predicting spectrogram in the second pass. However, predicting discrete units achieves 2.51x decoding speed-up compared to that case., Comment: ACL 2023 (main conference)

Reversible data hiding in encrypted domain(RDH-ED) can not only protect the privacy of 3D mesh models and embed additional data, but also recover original models and extract additional data losslessly. However, due to the insufficient use of model topology, the existing methods have not achieved satisfactory results in terms of embedding capacity. To further improve the capacity, a RDH-ED method is proposed based on the topology of the 3D mesh models, which divides the vertices into two parts: embedding set and prediction set. And after integer mapping, the embedding ability of the embedding set is calculated by the prediction set. It is then passed to the data hider for embedding additional data. Finally, the additional data and the original models can be extracted and recovered respectively by the receiver with the correct keys. Experiments declare that compared with the existing methods, this method can obtain the highest embedding capacity.

Direct speech-to-speech translation (S2ST) is among the most challenging problems in the translation paradigm due to the significant scarcity of S2ST data. While effort has been made to increase the data size from unlabeled speech by cascading pretrained speech recognition (ASR), machine translation (MT) and text-to-speech (TTS) models; unlabeled text has remained relatively under-utilized to improve S2ST. We propose an effective way to utilize the massive existing unlabeled text from different languages to create a large amount of S2ST data to improve S2ST performance by applying various acoustic effects to the generated synthetic data. Empirically our method outperforms the state of the art in Spanish-English translation by up to 2 BLEU. Significant gains by the proposed method are demonstrated in extremely low-resource settings for both Spanish-English and Russian-English translations.

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., Comment: \c{opyright} 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works

The amount of labeled data to train models for speech tasks is limited for most languages, however, the data scarcity is exacerbated for speech translation which requires labeled data covering two different languages. To address this issue, we study a simple and effective approach to build speech translation systems without labeled data by leveraging recent advances in unsupervised speech recognition, machine translation and speech synthesis, either in a pipeline approach, or to generate pseudo-labels for training end-to-end speech translation models. Furthermore, we present an unsupervised domain adaptation technique for pre-trained speech models which improves the performance of downstream unsupervised speech recognition, especially for low-resource settings. Experiments show that unsupervised speech-to-text translation outperforms the previous unsupervised state of the art by 3.2 BLEU on the Libri-Trans benchmark, on CoVoST 2, our best systems outperform the best supervised end-to-end models (without pre-training) from only two years ago by an average of 5.0 BLEU over five X-En directions. We also report competitive results on MuST-C and CVSS benchmarks.

We define a knot to be half ribbon if it is the cross-section of a ribbon 2-knot, and observe that ribbon implies half ribbon implies slice. We introduce the half ribbon genus of a knot K, the minimum genus of a ribbon knotted surface of which K is a cross-section. We compute this genus for all prime knots up to 12 crossings, and many 13-crossing knots. The same approach yields new computations of the doubly slice genus. We also introduce the half fusion number of a knot K, that measures the complexity of ribbon 2-knots of which K is a cross-section. We show that it is bounded from below by the Levine-Tristram signatures, and differs from the standard fusion number by an arbitrarily large amount., Comment: 13 pages, 1 figure, 1 table. Comments welcome. V2: typographical corrections

High-definition (HD) maps are essential in testing autonomous driving systems (ADSs). HD maps essentially determine the potential diversity of the testing scenarios. However, the current HD maps suffer from two main limitations: lack of junction diversity in the publicly available HD maps and cost-consuming to build a new HD map. Hence, in this paper, we propose, FEAT2MAP, to automatically generate concise HD maps with scenario diversity guarantees. FEAT2MAP focuses on junctions as they significantly influence scenario diversity, especially in urban road networks. FEAT2MAP first defines a set of features to characterize junctions. Then, FEAT2MAP extracts and samples concrete junction features from a list of input HD maps or user-defined requirements. Each junction feature generates a junction. Finally, FEAT2MAP builds a map by connecting the junctions in a grid layout. To demonstrate the effectiveness of FEAT2MAP, we conduct experiments with the public HD maps from SVL and the open-source ADS Apollo. The results show that FEAT2MAP can (1) generate new maps of reduced size while maintaining scenario diversity in terms of the code coverage and motion states of the ADS under test, and (2) generate new maps of increased scenario diversity by merging intersection features from multiple maps or taking user inputs., Comment: 7 pages, 7 figures

We describe a method to jointly pre-train speech and text in an encoder-decoder modeling framework for speech translation and recognition. The proposed method incorporates four self-supervised and supervised subtasks for cross modality learning. A self-supervised speech subtask leverages unlabelled speech data, and a (self-)supervised text to text subtask makes use of abundant text training data. Two auxiliary supervised speech tasks are included to unify speech and text modeling space. Our contribution lies in integrating linguistic information from the text corpus into the speech pre-training. Detailed analysis reveals learning interference among subtasks. Two pre-training configurations for speech translation and recognition, respectively, are presented to alleviate subtask interference. Our experiments show the proposed method can effectively fuse speech and text information into one model. It achieves between 1.7 and 2.3 BLEU improvement above the state of the art on the MuST-C speech translation dataset and comparable WERs to wav2vec 2.0 on the Librispeech speech recognition task., Comment: ACL 2022 main conference

Automated Driving Systems (ADSs) have seen rapid progress in recent years. To ensure the safety and reliability of these systems, extensive testings are being conducted before their future mass deployment. Testing the system on the road is the closest to real-world and desirable approach, but it is incredibly costly. Also, it is infeasible to cover rare corner cases using such real-world testing. Thus, a popular alternative is to evaluate an ADS's performance in some well-designed challenging scenarios, a.k.a. scenario-based testing. High-fidelity simulators have been widely used in this setting to maximize flexibility and convenience in testing what-if scenarios. Although many works have been proposed offering diverse frameworks/methods for testing specific systems, the comparisons and connections among these works are still missing. To bridge this gap, in this work, we provide a generic formulation of scenario-based testing in high-fidelity simulation and conduct a literature review on the existing works. We further compare them and present the open challenges as well as potential future research directions.

We present a direct simultaneous speech-to-speech translation (Simul-S2ST) model, Furthermore, the generation of translation is independent from intermediate text representations. Our approach leverages recent progress on direct speech-to-speech translation with discrete units, in which a sequence of discrete representations, instead of continuous spectrogram features, learned in an unsupervised manner, are predicted from the model and passed directly to a vocoder for speech synthesis on-the-fly. We also introduce the variational monotonic multihead attention (V-MMA), to handle the challenge of inefficient policy learning in speech simultaneous translation. The simultaneous policy then operates on source speech features and target discrete units. We carry out empirical studies to compare cascaded and direct approach on the Fisher Spanish-English and MuST-C English-Spanish datasets. Direct simultaneous model is shown to outperform the cascaded model by achieving a better tradeoff between translation quality and latency.

Speech-to-speech translation (S2ST) converts input speech to speech in another language. A challenge of delivering S2ST in real time is the accumulated delay between the translation and speech synthesis modules. While recently incremental text-to-speech (iTTS) models have shown large quality improvements, they typically require additional future text inputs to reach optimal performance. In this work, we minimize the initial waiting time of iTTS by adapting the upstream speech translator to generate high-quality pseudo lookahead for the speech synthesizer. After mitigating the initial delay, we demonstrate that the duration of synthesized speech also plays a crucial role on latency. We formalize this as a latency metric and then present a simple yet effective duration-scaling approach for latency reduction. Our approaches consistently reduce latency by 0.2-0.5 second without sacrificing speech translation quality., Comment: Accepted by Interspeech 2022

In this paper, we describe our end-to-end multilingual speech translation system submitted to the IWSLT 2021 evaluation campaign on the Multilingual Speech Translation shared task. Our system is built by leveraging transfer learning across modalities, tasks and languages. First, we leverage general-purpose multilingual modules pretrained with large amounts of unlabelled and labelled data. We further enable knowledge transfer from the text task to the speech task by training two tasks jointly. Finally, our multilingual model is finetuned on speech translation task-specific data to achieve the best translation results. Experimental results show our system outperforms the reported systems, including both end-to-end and cascaded based approaches, by a large margin. In some translation directions, our speech translation results evaluated on the public Multilingual TEDx test set are even comparable with the ones from a strong text-to-text translation system, which uses the oracle speech transcripts as input., Comment: Accepted by IWSLT 2021 as a system paper

Pretraining and multitask learning are widely used to improve the speech to text translation performance. In this study, we are interested in training a speech to text translation model along with an auxiliary text to text translation task. We conduct a detailed analysis to understand the impact of the auxiliary task on the primary task within the multitask learning framework. Our analysis confirms that multitask learning tends to generate similar decoder representations from different modalities and preserve more information from the pretrained text translation modules. We observe minimal negative transfer effect between the two tasks and sharing more parameters is helpful to transfer knowledge from the text task to the speech task. The analysis also reveals that the modality representation difference at the top decoder layers is still not negligible, and those layers are critical for the translation quality. Inspired by these findings, we propose three methods to improve translation quality. First, a parameter sharing and initialization strategy is proposed to enhance information sharing between the tasks. Second, a novel attention-based regularization is proposed for the encoders and pulls the representations from different modalities closer. Third, an online knowledge distillation is proposed to enhance the knowledge transfer from the text to the speech task. Our experiments show that the proposed approach improves translation performance by more than 2 BLEU over a strong baseline and achieves state-of-the-art results on the \textsc{MuST-C} English-German, English-French and English-Spanish language pairs., Comment: Accepted by ACL 2021

We present a direct speech-to-speech translation (S2ST) model that translates speech from one language to speech in another language without relying on intermediate text generation. We tackle the problem by first applying a self-supervised discrete speech encoder on the target speech and then training a sequence-to-sequence speech-to-unit translation (S2UT) model to predict the discrete representations of the target speech. When target text transcripts are available, we design a joint speech and text training framework that enables the model to generate dual modality output (speech and text) simultaneously in the same inference pass. Experiments on the Fisher Spanish-English dataset show that the proposed framework yields improvement of 6.7 BLEU compared with a baseline direct S2ST model that predicts spectrogram features. When trained without any text transcripts, our model performance is comparable to models that predict spectrograms and are trained with text supervision, showing the potential of our system for translation between unwritten languages. Audio samples are available at https://facebookresearch.github.io/speech_translation/direct_s2st_units/index.html ., Comment: Accepted to ACL 2022 (long paper)

Multi-head attention has each of the attention heads collect salient information from different parts of an input sequence, making it a powerful mechanism for sequence modeling. Multilingual and multi-domain learning are common scenarios for sequence modeling, where the key challenge is to maximize positive transfer and mitigate negative transfer across languages and domains. In this paper, we find that non-selective attention sharing is sub-optimal for achieving good generalization across all languages and domains. We further propose attention sharing strategies to facilitate parameter sharing and specialization in multilingual and multi-domain sequence modeling. Our approach automatically learns shared and specialized attention heads for different languages and domains to mitigate their interference. Evaluated in various tasks including speech recognition, text-to-text and speech-to-text translation, the proposed attention sharing strategies consistently bring gains to sequence models built upon multi-head attention. For speech-to-text translation, our approach yields an average of $+2.0$ BLEU over $13$ language directions in multilingual setting and $+2.0$ BLEU over $3$ domains in multi-domain setting.

In this work, we reanalyzed 11 years of spectral data from the \textit{Fermi} Large Area Telescope (\textit{Fermi}-LAT) of currently observed starburst galaxies (SBGs) and star-forming galaxies (SFGs). We used a one-zone model provided by \textbf{NAIMA} and the hadronic origin to explain the GeV observation data of the SBGs and SFGs. We found that a protonic distribution of a power-law form with an exponential cutoff can explain the spectra of most SBGs and SFGs. However, it cannot explain the spectral hardening components of NGC 1068 and NGC 4945 in the GeV energy band. Therefore, we considered the two-zone model to well explain these phenomena. We summarized the features of various model parameters, including the spectral index, cutoff energy, and protonic total energy. Similar to the evolution of supernova remnants (SNRs) in the Milky Way, we estimated the protonic acceleration limitation inside the SBGs to be likely in the order of 10$^{2}$ TeV using the one-zone model; this is close to those of SNRs in the Milky Way.

We present a simple yet effective approach to build multilingual speech-to-text (ST) translation by efficient transfer learning from pretrained speech encoder and text decoder. Our key finding is that a minimalistic LNA (LayerNorm and Attention) finetuning can achieve zero-shot crosslingual and cross-modality transfer ability by only finetuning less than 10% of the pretrained parameters. This enables effectively leveraging large pretrained models with low training cost. Using wav2vec 2.0 for acoustic modeling, and mBART for multilingual text generation, our approach advanced the new state-of-the-art for 34 translation directions (and surpassing cascaded ST for 23 of them) on large-scale multilingual ST benchmark CoVoST 2 (+6.4 BLEU on average across 15 En-X directions and +5.1 BLEU on average across 19 X-En directions). Our approach demonstrates strong zero-shot performance in a many-to-many multilingual model (+5.7 BLEU on average across 18 non-English directions), making it an appealing approach for attaining high-quality speech translation with improved parameter and data efficiency.

Attention-based sequence-to-sequence modeling provides a powerful and elegant solution for applications that need to map one sequence to a different sequence. Its success heavily relies on the availability of large amounts of training data. This presents a challenge for speech applications where labelled speech data is very expensive to obtain, such as automatic speech recognition (ASR) and speech translation (ST). In this study, we propose a general multi-task learning framework to leverage text data for ASR and ST tasks. Two auxiliary tasks, a denoising autoencoder task and machine translation task, are proposed to be co-trained with ASR and ST tasks respectively. We demonstrate that representing text input as phoneme sequences can reduce the difference between speech and text inputs, and enhance the knowledge transfer from text corpora to the speech to text tasks. Our experiments show that the proposed method achieves a relative 10~15% word error rate reduction on the English Librispeech task compared with our baseline, and improves the speech translation quality on the MuST-C tasks by 3.6~9.2 BLEU., Comment: Accepted for ICASSP2021

We introduce fairseq S2T, a fairseq extension for speech-to-text (S2T) modeling tasks such as end-to-end speech recognition and speech-to-text translation. It follows fairseq's careful design for scalability and extensibility. We provide end-to-end workflows from data pre-processing, model training to offline (online) inference. We implement state-of-the-art RNN-based, Transformer-based as well as Conformer-based models and open-source detailed training recipes. Fairseq's machine translation models and language models can be seamlessly integrated into S2T workflows for multi-task learning or transfer learning. Fairseq S2T documentation and examples are available at https://github.com/pytorch/fairseq/tree/master/examples/speech_to_text., Comment: Post-conference updates (accepted to AACL 2020 Demo)

The theory behind the electrical switching of antiferromagnets is premised on the existence of a well defined broken symmetry state that can be rotated to encode information. A spin glass is in many ways the antithesis of this state, characterized by an ergodic landscape of nearly degenerate magnetic configurations, choosing to freeze into a distribution of these in a manner that is seemingly bereft of information. In this study, we show that the coexistence of spin glass and antiferromagnetic order allows a novel mechanism to facilitate the switching of the antiferromagnet Fe$_{1/3+\delta}$NbS$_2$, which is rooted in the electrically-stimulated collective winding of the spin glass. The local texture of the spin glass opens an anisotropic channel of interaction that can be used to rotate the equilibrium orientation of the antiferromagnetic state. The use of a spin glass' collective dynamics to electrically manipulate antiferromagnetic spin textures has never been applied before, opening the field of antiferromagnetic spintronics to many more material platforms with complex magnetic textures., Comment: 7 pages, 4 Figures, supplement available on reasonable request

One of the main challenges for end-to-end speech translation is data scarcity. We leverage pseudo-labels generated from unlabeled audio by a cascade and an end-to-end speech translation model. This provides 8.3 and 5.7 BLEU gains over a strong semi-supervised baseline on the MuST-C English-French and English-German datasets, reaching state-of-the art performance. The effect of the quality of the pseudo-labels is investigated. Our approach is shown to be more effective than simply pre-training the encoder on the speech recognition task. Finally, we demonstrate the effectiveness of self-training by directly generating pseudo-labels with an end-to-end model instead of a cascade model., Comment: INTERSPEECH 2020

Translational distance-based knowledge graph embedding has shown progressive improvements on the link prediction task, from TransE to the latest state-of-the-art RotatE. However, N-1, 1-N and N-N predictions still remain challenging. In this work, we propose a novel translational distance-based approach for knowledge graph link prediction. The proposed method includes two-folds, first we extend the RotatE from 2D complex domain to high dimension space with orthogonal transforms to model relations for better modeling capacity. Second, the graph context is explicitly modeled via two directed context representations. These context representations are used as part of the distance scoring function to measure the plausibility of the triples during training and inference. The proposed approach effectively improves prediction accuracy on the difficult N-1, 1-N and N-N cases for knowledge graph link prediction task. The experimental results show that it achieves better performance on two benchmark data sets compared to the baseline RotatE, especially on data set (FB15k-237) with many high in-degree connection nodes., Comment: Accepted by ACL 2020

Machine reading comprehension(MRC) has attracted significant amounts of research attention recently, due to an increase of challenging reading comprehension datasets. In this paper, we aim to improve a MRC model's ability to determine whether a question has an answer in a given context (e.g. the recently proposed SQuAD 2.0 task). Our solution is a relation module that is adaptable to any MRC model. The relation module consists of both semantic extraction and relational information. We first extract high level semantics as objects from both question and context with multi-head self-attentive pooling. These semantic objects are then passed to a relation network, which generates relationship scores for each object pair in a sentence. These scores are used to determine whether a question is non-answerable. We test the relation module on the SQuAD 2.0 dataset using both BiDAF and BERT models as baseline readers. We obtain 1.8% gain of F1 on top of the BiDAF reader, and 1.0% on top of the BERT base model. These results show the effectiveness of our relation module on MRC

Recent years have seen great success in the use of neural seq2seq models on the text-to-SQL task. However, little work has paid attention to how these models generalize to realistic unseen data, which naturally raises a question: does this impressive performance signify a perfect generalization model, or are there still some limitations? In this paper, we first diagnose the bottleneck of text-to-SQL task by providing a new testbed, in which we observe that existing models present poor generalization ability on rarely-seen data. The above analysis encourages us to design a simple but effective auxiliary task, which serves as a supportive model as well as a regularization term to the generation task to increase the models generalization. Experimentally, We evaluate our models on a large text-to-SQL dataset WikiSQL. Compared to a strong baseline coarse-to-fine model, our models improve over the baseline by more than 3% absolute in accuracy on the whole dataset. More interestingly, on a zero-shot subset test of WikiSQL, our models achieve 5% absolute accuracy gain over the baseline, clearly demonstrating its superior generalizability.

Multi-hop reading comprehension (RC) across documents poses new challenge over single-document RC because it requires reasoning over multiple documents to reach the final answer. In this paper, we propose a new model to tackle the multi-hop RC problem. We introduce a heterogeneous graph with different types of nodes and edges, which is named as Heterogeneous Document-Entity (HDE) graph. The advantage of HDE graph is that it contains different granularity levels of information including candidates, documents and entities in specific document contexts. Our proposed model can do reasoning over the HDE graph with nodes representation initialized with co-attention and self-attention based context encoders. We employ Graph Neural Networks (GNN) based message passing algorithms to accumulate evidences on the proposed HDE graph. Evaluated on the blind test set of the Qangaroo WikiHop data set, our HDE graph based single model delivers competitive result, and the ensemble model achieves the state-of-the-art performance., Comment: To appear in ACL 2019

The I4U consortium was established to facilitate a joint entry to NIST speaker recognition evaluations (SRE). The latest edition of such joint submission was in SRE 2018, in which the I4U submission was among the best-performing systems. SRE'18 also marks the 10-year anniversary of I4U consortium into NIST SRE series of evaluation. The primary objective of the current paper is to summarize the results and lessons learned based on the twelve sub-systems and their fusion submitted to SRE'18. It is also our intention to present a shared view on the advancements, progresses, and major paradigm shifts that we have witnessed as an SRE participant in the past decade from SRE'08 to SRE'18. In this regard, we have seen, among others, a paradigm shift from supervector representation to deep speaker embedding, and a switch of research challenge from channel compensation to domain adaptation., Comment: 5 pages

The interplay of symmetry and quenched disorder leads to some of the most fundamentally interesting and technologically important properties of correlated materials. It also poses the most vexing of theoretical challenges. Nowhere is this more apparent than in the study of spin glasses. A spin glass is characterized by an ergodic landscape of states - an innumerable number of possibilities that are only weakly distinguished energetically, if at all. We show in the material Fe$_x$NbS$_2$, this landscape of states can be biased by coexisitng antiferromagnetic order. This process leads to a phenomenon of broad technological importance: giant, tunable exchange bias. We observe exchange biases that exceed those of conventional materials by more than two orders of magnitude. This work illustrates a novel route to giant exchange bias by leveraging the interplay of frustration and disorder in exotic materials., Comment: 18 pages, 4 figures

Speech emotion recognition (SER) has attracted great attention in recent years due to the high demand for emotionally intelligent speech interfaces. Deriving speaker-invariant representations for speech emotion recognition is crucial. In this paper, we propose to apply adversarial training to SER to learn speaker-invariant representations. Our model consists of three parts: a representation learning sub-network with time-delay neural network (TDNN) and LSTM with statistical pooling, an emotion classification network and a speaker classification network. Both the emotion and speaker classification network take the output of the representation learning network as input. Two training strategies are employed: one based on domain adversarial training (DAT) and the other one based on cross-gradient training (CGT). Besides the conventional data set, we also evaluate our proposed models on a much larger publicly available emotion data set with 250 speakers. Evaluation results show that on IEMOCAP, DAT and CGT provides 5.6% and 7.4% improvement respectively, over a baseline system without speaker-invariant representation learning on 5-fold cross validation. On the larger emotion data set, while CGT fails to yield better results than baseline, DAT can still provide 9.8% relative improvement on a standalone test set.

This paper aims to improve the widely used deep speaker embedding x-vector model. We propose the following improvements: (1) a hybrid neural network structure using both time delay neural network (TDNN) and long short-term memory neural networks (LSTM) to generate complementary speaker information at different levels; (2) a multi-level pooling strategy to collect speaker information from both TDNN and LSTM layers; (3) a regularization scheme on the speaker embedding extraction layer to make the extracted embeddings suitable for the following fusion step. The synergy of these improvements are shown on the NIST SRE 2016 eval test (with a 19% EER reduction) and SRE 2018 dev test (with a 9% EER reduction), as well as more than 10% DCF scores reduction on these two test sets over the x-vector baseline., Comment: Accepted by ICASSP2019

Knowledge graph embedding has been an active research topic for knowledge base completion, with progressive improvement from the initial TransE, TransH, DistMult et al to the current state-of-the-art ConvE. ConvE uses 2D convolution over embeddings and multiple layers of nonlinear features to model knowledge graphs. The model can be efficiently trained and scalable to large knowledge graphs. However, there is no structure enforcement in the embedding space of ConvE. The recent graph convolutional network (GCN) provides another way of learning graph node embedding by successfully utilizing graph connectivity structure. In this work, we propose a novel end-to-end Structure-Aware Convolutional Network (SACN) that takes the benefit of GCN and ConvE together. SACN consists of an encoder of a weighted graph convolutional network (WGCN), and a decoder of a convolutional network called Conv-TransE. WGCN utilizes knowledge graph node structure, node attributes and edge relation types. It has learnable weights that adapt the amount of information from neighbors used in local aggregation, leading to more accurate embeddings of graph nodes. Node attributes in the graph are represented as additional nodes in the WGCN. The decoder Conv-TransE enables the state-of-the-art ConvE to be translational between entities and relations while keeps the same link prediction performance as ConvE. We demonstrate the effectiveness of the proposed SACN on standard FB15k-237 and WN18RR datasets, and it gives about 10% relative improvement over the state-of-the-art ConvE in terms of HITS@1, HITS@3 and HITS@10., Comment: The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI 2019)

In August 2010, the Smithsonian Science Education Center (SSEC) received a grant of more than $25 million from the U.S. Department of Education's Investing in Innovation (i3) program for a five-year study to validate its Leadership Assistance for Science Education Reform (LASER) model in three very diverse regions of the United States: rural North Carolina, northern New Mexico, and the Houston Independent School District (HISD). A random sample of schools in the three regions took the PASS Open-Ended and Performance Task assessment for the first time in Spring 2012, (end of first posttest year) and again in Spring 2013 and Spring 2014 (second and third posttest years, respectively). Students in the elementary cohort (currently 5th graders in 2013-2014) responded to two Open-Ended (OE) and six Performance Task (PT) items, while students in the middle school cohort (currently 8th graders in 2013-2014) responded to six OE and six PT items. Results are presented in this report.

In August 2010, the Smithsonian Science Education Center (SSEC) received a grant of more than $25 million from the U.S. Department of Education's Investing in Innovation (i3) program for a five-year study to validate its Leadership Assistance for Science Education Reform (LASER) model in three very diverse regions of the United States: rural North Carolina, northern New Mexico, and the Houston Independent School District (HISD). State assessment results from the Spring 2014 (third posttest for the Elementary and the Middle School Cohort) administrations are available and are reported in this report. As the state assessments for each region are different, results are reported by region only (and not in the aggregate across regions). Outcomes for the Houston Independent School District (HISD) are reported first (State of Texas Assessment of Academic Readiness (STAAR) and Stanford), followed by the New Mexico (Standards Based Assessment (SBA)) and North Carolina (End-of-Grade (EOG)) regions. A summary of the Key Findings for each set of analyses is presented at the beginning of each report, followed by information on the samples included, baseline equivalence between the Phase 1 and Phase 2 groups, and the detailed outcomes by grade level (i.e., elementary cohort and middle school cohort) and subgroup.

In August 2010, the Smithsonian Science Education Center (SSEC) received a grant of more than $25 million from the U.S. Department of Education's Investing in Innovation (i3) program for a five-year study to validate its Leadership Assistance for Science Education Reform (LASER) model in three very diverse regions of the United States: rural North Carolina, northern New Mexico, and the Houston Independent School District (HISD). PASS student attitudes survey data results from the Fall 2011 (baseline or pre-intervention) and Spring 2014 (third posttest for the Elementary and the Middle School Cohort) administrations are available and are reported in this report. There were a total of 14 questions on the PASS that addressed classroom achievement in core subjects (reading, math, and science) and different aspects of science engagement and science activities in the school.

In August 2010, the Smithsonian Science Education Center (SSEC) received a grant of more than $25 million from the U.S. Department of Education's Investing in Innovation (i3) program for a five-year study to validate its Leadership Assistance for Science Education Reform (LASER) model in three very diverse regions of the United States: rural North Carolina, northern New Mexico, and the Houston Independent School District (HISD). There were a total of 29 multiple choice questions on both the Fall 2011 and Spring 2014 forms of the PASS (PASS-MC) addressing five broad science content standard categories for the elementary cohort and six broad science content standard categories for the middle school cohort. Only students who answered at least one multiple choice achievement question at both time points were included in the analyses for each respective area of analysis. Results for All Regions combined are presented first, followed by the outcomes for the Houston Independent School District (HISD), the New Mexico region, and the North Carolina region. A summary of the Key Findings for each set of analyses is presented at the beginning of each report, followed by information on the samples included, baseline equivalence between the Phase 1 and Phase 2 groups, and the detailed outcomes by grade level (i.e., elementary cohort and middle school cohort) and subgroup.

Neutron penumbral imaging technique has been successfully used as the diagnosis method in Inertial Confined Fusion. To help the design of the imaging systems in the future in CHINA. We construct the Monte carlo imaging system by Geant4. Use the point spread function from the simulation and decode algorithm (Lucy-Rechardson algorithm) we got the recovery image., Comment: 5 pages,13 figures

We present an improved calculation on the pionic twist-3 distribution amplitudes $\phi^{\pi}_{p}$ and $\phi^{\pi}_{\sigma}$, which are studied within the QCD sum rules. By adding all the uncertainties in quadrature, it is found that $<\xi^2_p>=0.248^{+0.076}_{-0.052}$, $<\xi^4_p>=0.262^{+0.080}_{-0.055}$, $<\xi^2_\sigma>=0.102^{+0.035}_{-0.025}$ and $<\xi^4_\sigma>=0.094^{+0.028}_{-0.020}$. Furthermore, with the help of these moments, we construct a model for the twist-3 wave functions $\psi^{\pi}_{p,\sigma}(x,\mathbf{k}_\bot)$, which have better end-point behavior and are helpful for perturbative QCD approach. The obtained twist-3 distribution amplitudes are adopted to calculate the $B\to\pi$ transition form factor $f^+_{B\pi}$ within the QCD light-cone sum rules up to next-to-leading order. By suitable choice of the parameters, we obtain a consistent $f^+_{B\pi}$ with those obtained in the literature., Comment: 9 pages, 6 figures and 6 tables. To be published in Phys.Rev.D

We study the domain growth of the one-dimensional kinetic Ising model under the competing influence of Glauber dynamics at temperature T and Kawasaki dynamics with a configuration-independent rate. The scaling of the structure factor is shown to have the form for nonconserved dynamics with the corrections arising from the spin-exchange process, i.e., $S(k,t)=Lg_0(kL,t/\tau )+g_1(kL,t/\tau)+... $, and the corresponding scaling functions are calculated analytically. A correction to the Porod law at zero temperature is also given., Comment: 4 pages, RevTex