arXiv subjects

Bowen Shi

Publications and source records attributed to Bowen Shi.

At least 73 records · Page 4Linked to original sources

arXiv2023

Scaling Speech Technology to 1,000+ Languages

Expanding the language coverage of speech technology has the potential to improve access to information for many more people. However, current speech technology is restricted to about one hundred languages which is a small fraction of the over 7,000 languages spoken around the world. The Massively Multilingual Speech (MMS) project increases the number of supported languages by 10-40x, depending on the task. The main ingredients are a new dataset based on readings of publicly available religious texts and effectively leveraging self-supervised learning. We built pre-trained wav2vec 2.0 models covering 1,406 languages, a single multilingual automatic speech recognition model for 1,107 languages, speech synthesis models for the same number of languages, as well as a language identification model for 4,017 languages. Experiments show that our multilingual speech recognition model more than halves the word error rate of Whisper on 54 languages of the FLEURS benchmark while being trained on a small fraction of the labeled data.

cs.CL↗

arXiv2023

Highly-confined and tunable plasmonics based on two-dimensional solid-state defect lattices

Plasmons, collective excitations of electrons in solids, are associated with strongly confined electromagnetic fields, with wavelengths far below the wavelength of photons in free space. This strong confinement promises the realization of optoelectronic devices that could bridge the size difference between photonic and electronic devices. However, despite decades of research in plasmonics, many applications remain limited by plasmonic losses, thus motivating a search for new engineered plasmonic materials with lower losses. A promising pathway for low-loss plasmonic materials is the engineering of materials with flat and energetically isolated metallic bands, which can strongly limit phonon-assisted optical losses, a major contributor to short plasmonic lifetimes. Such electronic band structures may be created by judiciously introducing an ordered lattice of defects in an insulating host material. Here, we explore this approach, presenting several low-loss, highly-confined, and tunable plasmonic materials based on arrays of carbon substitutions in hexagonal boron nitride (hBN) monolayers. From our first-principles calculations based on density functional theory (DFT), we find plasmonic structures with mid-infrared plasmons featuring very high confinements ($λ_{\text{vacuum}}/λ_{\text{plasmon}}$ exceeding 2000) and quality factors in excess of 1000. We provide a systematic explanation of how crystal structure, electronic bandwidth, and many-body effects affect the plasmonic dispersions and losses of these materials. The results are thus of relevance to low-loss plasmon engineering in other flat band systems.

physics.optics↗

arXiv2023

Hierarchical Graph Networks for 3D Human Pose Estimation

Recent 2D-to-3D human pose estimation works tend to utilize the graph structure formed by the topology of the human skeleton. However, we argue that this skeletal topology is too sparse to reflect the body structure and suffer from serious 2D-to-3D ambiguity problem. To overcome these weaknesses, we propose a novel graph convolution network architecture, Hierarchical Graph Networks (HGN). It is based on denser graph topology generated by our multi-scale graph structure building strategy, thus providing more delicate geometric information. The proposed architecture contains three sparse-to-fine representation subnetworks organized in parallel, in which multi-scale graph-structured features are processed and exchange information through a novel feature fusion strategy, leading to rich hierarchical representations. We also introduce a 3D coarse mesh constraint to further boost detail-related feature learning. Extensive experiments demonstrate that our HGN achieves the state-of-the art performance with reduced network parameters. Code is released at https://github.com/qingshi9974/BMVC2021-Hierarchical-Graph-Networks-for-3D-Human-Pose-Estimation.

cs.CV↗

arXiv2023

Comparative layer-wise analysis of self-supervised speech models

Many self-supervised speech models, varying in their pre-training objective, input modality, and pre-training data, have been proposed in the last few years. Despite impressive successes on downstream tasks, we still have a limited understanding of the properties encoded by the models and the differences across models. In this work, we examine the intermediate representations for a variety of recent models. Specifically, we measure acoustic, phonetic, and word-level properties encoded in individual layers, using a lightweight analysis tool based on canonical correlation analysis (CCA). We find that these properties evolve across layers differently depending on the model, and the variations relate to the choice of pre-training objective. We further investigate the utility of our analyses for downstream tasks by comparing the property trends with performance on speech recognition and spoken language understanding tasks. We discover that CCA trends provide reliable guidance to choose layers of interest for downstream tasks and that single-layer performance often matches or improves upon using all layers, suggesting implications for more efficient use of pre-trained models.

cs.CL↗

arXiv2023

MuAViC: A Multilingual Audio-Visual Corpus for Robust Speech Recognition and Robust Speech-to-Text Translation

We introduce MuAViC, a multilingual audio-visual corpus for robust speech recognition and robust speech-to-text translation providing 1200 hours of audio-visual speech in 9 languages. It is fully transcribed and covers 6 English-to-X translation as well as 6 X-to-English translation directions. To the best of our knowledge, this is the first open benchmark for audio-visual speech-to-text translation and the largest open benchmark for multilingual audio-visual speech recognition. Our baseline results show that MuAViC is effective for building noise-robust speech recognition and translation models. We make the corpus available at https://github.com/facebookresearch/muavic.

cs.CL↗

arXiv2023

Pose-Oriented Transformer with Uncertainty-Guided Refinement for 2D-to-3D Human Pose Estimation

There has been a recent surge of interest in introducing transformers to 3D human pose estimation (HPE) due to their powerful capabilities in modeling long-term dependencies. However, existing transformer-based methods treat body joints as equally important inputs and ignore the prior knowledge of human skeleton topology in the self-attention mechanism. To tackle this issue, in this paper, we propose a Pose-Oriented Transformer (POT) with uncertainty guided refinement for 3D HPE. Specifically, we first develop novel pose-oriented self-attention mechanism and distance-related position embedding for POT to explicitly exploit the human skeleton topology. The pose-oriented self-attention mechanism explicitly models the topological interactions between body joints, whereas the distance-related position embedding encodes the distance of joints to the root joint to distinguish groups of joints with different difficulties in regression. Furthermore, we present an Uncertainty-Guided Refinement Network (UGRN) to refine pose predictions from POT, especially for the difficult joints, by considering the estimated uncertainty of each joint with uncertainty-guided sampling strategy and self-attention mechanism. Extensive experiments demonstrate that our method significantly outperforms the state-of-the-art methods with reduced model parameters on 3D HPE benchmarks such as Human3.6M and MPI-INF-3DHP

cs.CV↗

arXiv2023

Visual Story Generation Based on Emotion and Keywords

Automated visual story generation aims to produce stories with corresponding illustrations that exhibit coherence, progression, and adherence to characters' emotional development. This work proposes a story generation pipeline to co-create visual stories with the users. The pipeline allows the user to control events and emotions on the generated content. The pipeline includes two parts: narrative and image generation. For narrative generation, the system generates the next sentence using user-specified keywords and emotion labels. For image generation, diffusion models are used to create a visually appealing image corresponding to each generated sentence. Further, object recognition is applied to the generated images to allow objects in these images to be mentioned in future story development.

cs.AI↗

arXiv2022

ReVISE: Self-Supervised Speech Resynthesis with Visual Input for Universal and Generalized Speech Enhancement

Prior works on improving speech quality with visual input typically study each type of auditory distortion separately (e.g., separation, inpainting, video-to-speech) and present tailored algorithms. This paper proposes to unify these subjects and study Generalized Speech Enhancement, where the goal is not to reconstruct the exact reference clean signal, but to focus on improving certain aspects of speech. In particular, this paper concerns intelligibility, quality, and video synchronization. We cast the problem as audio-visual speech resynthesis, which is composed of two steps: pseudo audio-visual speech recognition (P-AVSR) and pseudo text-to-speech synthesis (P-TTS). P-AVSR and P-TTS are connected by discrete units derived from a self-supervised speech model. Moreover, we utilize self-supervised audio-visual speech model to initialize P-AVSR. The proposed model is coined ReVISE. ReVISE is the first high-quality model for in-the-wild video-to-speech synthesis and achieves superior performance on all LRS3 audio-visual enhancement tasks with a single model. To demonstrates its applicability in the real world, ReVISE is also evaluated on EasyCom, an audio-visual benchmark collected under challenging acoustic conditions with only 1.6 hours of training data. Similarly, ReVISE greatly suppresses noise and improves quality. Project page: https://wnhsu.github.io/ReVISE.

eess.AS↗

arXiv2022

u-HuBERT: Unified Mixed-Modal Speech Pretraining And Zero-Shot Transfer to Unlabeled Modality

While audio-visual speech models can yield superior performance and robustness compared to audio-only models, their development and adoption are hindered by the lack of labeled and unlabeled audio-visual data and the cost to deploy one model per modality. In this paper, we present u-HuBERT, a self-supervised pre-training framework that can leverage both multimodal and unimodal speech with a unified masked cluster prediction objective. By utilizing modality dropout during pre-training, we demonstrate that a single fine-tuned model can achieve performance on par or better than the state-of-the-art modality-specific models. Moreover, our model fine-tuned only on audio can perform well with audio-visual and visual speech input, achieving zero-shot modality generalization for multiple speech processing tasks. In particular, our single model yields 1.2%/1.4%/27.2% speech recognition word error rate on LRS3 with audio-visual/audio/visual input. Codes and models are available at https://github.com/facebookresearch/av_hubert

cs.CL↗

arXiv2022

Open-Domain Sign Language Translation Learned from Online Video

Existing work on sign language translation - that is, translation from sign language videos into sentences in a written language - has focused mainly on (1) data collected in a controlled environment or (2) data in a specific domain, which limits the applicability to real-world settings. In this paper, we introduce OpenASL, a large-scale American Sign Language (ASL) - English dataset collected from online video sites (e.g., YouTube). OpenASL contains 288 hours of ASL videos in multiple domains from over 200 signers and is the largest publicly available ASL translation dataset to date. To tackle the challenges of sign language translation in realistic settings and without glosses, we propose a set of techniques including sign search as a pretext task for pre-training and fusion of mouthing and handshape features. The proposed techniques produce consistent and large improvements in translation quality, over baseline models based on prior work. Our data and code are publicly available at https://github.com/chevalierNoir/OpenASL

cs.CV↗

arXiv2022

Robust Self-Supervised Audio-Visual Speech Recognition

Audio-based automatic speech recognition (ASR) degrades significantly in noisy environments and is particularly vulnerable to interfering speech, as the model cannot determine which speaker to transcribe. Audio-visual speech recognition (AVSR) systems improve robustness by complementing the audio stream with the visual information that is invariant to noise and helps the model focus on the desired speaker. However, previous AVSR work focused solely on the supervised learning setup; hence the progress was hindered by the amount of labeled data available. In this work, we present a self-supervised AVSR framework built upon Audio-Visual HuBERT (AV-HuBERT), a state-of-the-art audio-visual speech representation learning model. On the largest available AVSR benchmark dataset LRS3, our approach outperforms prior state-of-the-art by ~50% (28.0% vs. 14.1%) using less than 10% of labeled data (433hr vs. 30hr) in the presence of babble noise, while reducing the WER of an audio-based model by over 75% (25.8% vs. 5.8%) on average.

cs.SD↗

arXiv2022

Learning Lip-Based Audio-Visual Speaker Embeddings with AV-HuBERT

This paper investigates self-supervised pre-training for audio-visual speaker representation learning where a visual stream showing the speaker's mouth area is used alongside speech as inputs. Our study focuses on the Audio-Visual Hidden Unit BERT (AV-HuBERT) approach, a recently developed general-purpose audio-visual speech pre-training framework. We conducted extensive experiments probing the effectiveness of pre-training and visual modality. Experimental results suggest that AV-HuBERT generalizes decently to speaker related downstream tasks, improving label efficiency by roughly ten fold for both audio-only and audio-visual speaker verification. We also show that incorporating visual information, even just the lip area, greatly improves the performance and noise robustness, reducing EER by 38% in the clean condition and 75% in noisy conditions.

eess.AS↗

arXiv2022

Modular Commutators in Conformal Field Theory

The modular commutator is a recently discovered multipartite entanglement measure that quantifies the chirality of the underlying many-body quantum state. In this Letter, we derive a universal expression for the modular commutator in conformal field theories in $1+1$ dimensions and discuss its salient features. We show that the modular commutator depends only on the chiral central charge and the conformal cross ratio. We test this formula for a gapped $(2+1)$-dimensional system with a chiral edge, i.e., the quantum Hall state, and observe excellent agreement with numerical simulations. Furthermore, we propose a geometric dual for the modular commutator in certain preferred states of the AdS/CFT correspondence. For these states, we argue that the modular commutator can be obtained from a set of crossing angles between intersecting Ryu-Takayanagi surfaces.

cond-mat.str-el↗

arXiv2022

Searching for fingerspelled content in American Sign Language

Natural language processing for sign language video - including tasks like recognition, translation, and search - is crucial for making artificial intelligence technologies accessible to deaf individuals, and is gaining research interest in recent years. In this paper, we address the problem of searching for fingerspelled key-words or key phrases in raw sign language videos. This is an important task since significant content in sign language is often conveyed via fingerspelling, and to our knowledge the task has not been studied before. We propose an end-to-end model for this task, FSS-Net, that jointly detects fingerspelling and matches it to a text sequence. Our experiments, done on a large public dataset of ASL fingerspelling in the wild, show the importance of fingerspelling detection as a component of a search and retrieval model. Our model significantly outperforms baseline methods adapted from prior work on related tasks

cs.CV↗

arXiv2022

Learning Audio-Visual Speech Representation by Masked Multimodal Cluster Prediction

Video recordings of speech contain correlated audio and visual information, providing a strong signal for speech representation learning from the speaker's lip movements and the produced sound. We introduce Audio-Visual Hidden Unit BERT (AV-HuBERT), a self-supervised representation learning framework for audio-visual speech, which masks multi-stream video input and predicts automatically discovered and iteratively refined multimodal hidden units. AV-HuBERT learns powerful audio-visual speech representation benefiting both lip-reading and automatic speech recognition. On the largest public lip-reading benchmark LRS3 (433 hours), AV-HuBERT achieves 32.5% WER with only 30 hours of labeled data, outperforming the former state-of-the-art approach (33.6%) trained with a thousand times more transcribed video data (31K hours). The lip-reading WER is further reduced to 26.9% when using all 433 hours of labeled data from LRS3 and combined with self-training. Using our audio-visual representation on the same benchmark for audio-only speech recognition leads to a 40% relative WER reduction over the state-of-the-art performance (1.3% vs 2.3%). Our code and models are available at https://github.com/facebookresearch/av_hubert

eess.AS↗

arXiv2021

Modular commutator in gapped quantum many-body systems

In arXiv:2110.06932, we argued that the chiral central charge -- a topologically protected quantity characterizing the edge theory of a gapped (2+1)-dimensional system -- can be extracted from the bulk by using an order parameter called the modular commutator. In this paper, we reveal general properties of the modular commutator and strengthen its relationship with the chiral central charge. First, we identify connections between the modular commutator and conditional mutual information, time reversal, and modular flow. Second, we prove, within the framework of the entanglement bootstrap program, that two topologically ordered media connected by a gapped domain wall must have the same modular commutator in their respective bulk. Third, we numerically calculate the value of the modular commutator for a bosonic lattice Laughlin state for finite sizes and extrapolate to the infinite-volume limit. The result of this extrapolation is consistent with the proposed formula up to an error of about 0.7%.

quant-ph↗

arXiv2021

Chiral central charge from a single bulk wave function

A $(2+1)$-dimensional gapped quantum many-body system can have a topologically protected energy current at its edge. The magnitude of this current is determined entirely by the temperature and the chiral central charge, a quantity associated with the effective field theory of the edge. We derive a formula for the chiral central charge that, akin to the topological entanglement entropy, is completely determined by the many-body ground state wave function in the bulk. According to our formula, nonzero chiral central charge gives rise to a topological obstruction that prevents the ground state wave function from being real-valued in any local product basis.

quant-ph↗

arXiv2021

Entanglement negativity at the critical point of measurement-driven transition

We study the entanglement behavior of a random unitary circuit punctuated by projective measurements at the measurement-driven phase transition in one spatial dimension. We numerically study the logarithmic entanglement negativity of two disjoint intervals and find that it scales as a power of the cross-ratio. We investigate two systems: (1) Clifford circuits with projective measurements, and (2) Haar random local unitary circuit with projective measurements. Remarkably, we identify a power-law behavior of entanglement negativity at the critical point. Previous results of entanglement entropy and mutual information point to an emergent conformal invariance of the measurement-driven transition. Our result suggests that the critical behavior of the measurement-driven transition is distinct from the ground state behavior of any \emph{unitary} conformal field theory.

cond-mat.stat-mech↗