We release a multi-accent dataset and propose speech-programming and gradient reversal classifier to improve the generalization.Abstract: Speech-to-SQL (S2SQL) aims to convert spoken questions into SQL queries given relational databases, which has been traditionally implemented in a cascaded manner while facing the following challenges: 1) model training is faced with the major issue of data scarcity, where limited parallel data is available; and 2) the systems should be robust enough to handle diverse out-of-domain speech samples that differ from the source data. In this work, we propose the direct generalizable speech-to-SQL parsing model Wav2SQL which avoids error compounding across cascaded systems. Specifically, 1) to accelerate speech-driven SQL parsing research in the community, we release a large-scale and multi-accent dataset MASpider; 2) leveraging the recent progress in the large-scale pre-training, we show that it alleviates the data scarcity issue and allow for direct speech-to-SQL parsing; and 3) we include the speech re-programming and gradient reversal classifier techniques to reduce acoustic variance and learned style-agnostic representation, improving generalization to unseen out-of-domain custom data. Experimental results demonstrate that Wav2SQL avoids error compounding and achieves state-of-the-art results by up to 4.7% accuracy improvement over the baseline.

Speech-to-singing voice conversion (STS) task always suffers from data scarcity, because it requires paired speech and singing data. Compounding this issue are the challenges of content-pitch alignment and the suboptimal quality of generated outputs, presenting significant hurdles in STS research. This paper presents SVPT, an STS approach boosted by a self-supervised singing voice pre-training model.We leverage spoken language model techniques to tackle the rhythm alignment problem and the in-context learning capability to achieve zero-shot conversion. We adopt discrete-unit random resampling and pitch corruption strategies, enabling training with unpaired singing data and thus mitigating the issue of data scarcity. SVPT also serves as an effective backbone for singing voice synthesis (SVS), offering insights into scaling up SVS models. Experimental results indicate that SVPT delivers notable improvements in both STS and SVS endeavors. Audio samples are available at https://speech2sing.github.io.

Direct speech-to-speech translation achieves high-quality results through the introduction of discrete units obtained from self-supervised learning. However, talking head translation, converting audio-visual speech (i.e., talking head video) from one language into another, still confronts several challenges compared to audio speech: (1) Existing methods invariably rely on cascading, synthesizing via both audio and text, resulting in delays and cascading errors. (2) Talking head translation has a limited set of reference frames. If the generated translation exceeds the length of the original speech, the video sequence needs to be supplemented by repeating frames, leading to jarring video transitions. In this work, we propose a model for talking head translation, TransFace, which can directly translate audio-visual speech into audio-visual speech in other languages. It consists of a speech-to-unit translation model to convert audio speech into discrete units and a unit-based audio-visual speech synthesizer, Unit2Lip, to re-synthesize synchronized audio-visual speech from discrete units in parallel. Furthermore, we introduce a Bounded Duration Predictor, ensuring isometric talking head translation and preventing duplicate reference frames. Experiments demonstrate that Unit2Lip significantly improves synchronization and boosts inference speed by a factor of 4.35 on LRS2. Additionally, TransFace achieves impressive BLEU scores of 61.93 and 47.55 for Es-En and Fr-En on LRS3-T and 100% isochronous translations. The samples are available at https://transface-demo.github.io .

Recent singing-voice-synthesis (SVS) methods have achieved remarkable audio quality and naturalness, yet they lack the capability to control the style attributes of the synthesized singing explicitly. We propose Prompt-Singer, the first SVS method that enables attribute controlling on singer gender, vocal range and volume with natural language. We adopt a model architecture based on a decoder-only transformer with a multi-scale hierarchy, and design a range-melody decoupled pitch representation that enables text-conditioned vocal range control while keeping melodic accuracy. Furthermore, we explore various experiment settings, including different types of text representations, text encoder fine-tuning, and introducing speech data to alleviate data scarcity, aiming to facilitate further research. Experiments show that our model achieves favorable controlling ability and audio quality. Audio samples are available at http://prompt-singer.github.io .

The development of multimodal models has significantly advanced multimodal sentiment analysis and emotion recognition. However, in real-world applications, the presence of various missing modality cases often leads to a degradation in the model’s performance. In this work, we propose a novel multimodal Transformer framework using prompt learning to address the issue of missing modalities. Our method introduces three types of prompts: generative prompts, missing-signal prompts, and missing-type prompts. These prompts enable the generation of missing modality features and facilitate the learning of intra- and inter-modality information. Through prompt learning, we achieve a substantial reduction in the number of trainable parameters. Our proposed method outperforms other methods significantly across all evaluation metrics. Extensive experiments and ablation studies are conducted to demonstrate the effectiveness and robustness of our method, showcasing its ability to effectively handle missing modalities. Codes are available at https://github.com/zrguo/MPLMM.

Recently, instruction-following audio-language models have received broad attention for human-audio interaction. However, the absence of benchmarks capable of evaluating audio-centric interaction capabilities has impeded advancements in this field. Previous models primarily focus on assessing different fundamental tasks, such as automatic speech recognition, and lack an assessment of the open-ended generative capabilities centered around audio. Thus, it is challenging to track the progression in the Large Audio-Language Models (LALMs) domain and to provide guidance for future improvement.In this paper, we introduce AIR-Bench (Audio InstRuction Benchmark), the first benchmark designed to evaluate the ability of LALMs to understand various types of audio signals (including human speech, natural sounds, and music), and furthermore, to interact with humans in the textual format. AIR-Bench encompasses two dimensions: foundation and chat benchmarks. The former consists of 19 tasks with approximately 19k single-choice questions, intending to inspect the basic single-task ability of LALMs. The latter one contains 2k instances of open-ended question-and-answer data, directly assessing the comprehension of the model on complex audio and its capacity to follow instructions. Both benchmarks require the model to generate hypotheses directly. We design a unified framework that leverages advanced language models, such as GPT-4, to evaluate the scores of generated hypotheses given the meta-information of the audio. Experimental results demonstrate a high level of consistency between GPT-4-based evaluation and human evaluation. By revealing the limitations of existing LALMs through evaluation results, AIR-Bench can provide insights into the direction of future research. Dataset and evaluation code are available at https://github.com/OFA-Sys/AIR-Bench.

Transformer-based methods have gone mainstream in multimodal sequential learning. The intra and inter modality interactions are captured by the query-key associations of multi-head attention. In this way, the calculated multimodal contexts (attentional results) are expected to be relevant to the query modality. However, in existing literature, the alignment degree between different calculated attentional results of the same query are under-explored. Based on this concern, we propose a new constrained scheme called Multimodal Contextual Contrast (MCC), which could align the multiple attentional results from both local and global perspectives, making the information capture more efficient. Concretely, the calculated attentional results of different modalities are mapped into a common feature space, those attentional vectors with the same query are considered as a positive group and the remaining sets are negative. From local perspective, we sample the negative groups for a positive group by randomly changing the sequential step of one specific context and keeping the other stay the same. From coarse global perspective, we divide all the contextual groups into two sets (i.e., aligned and unaligned), making the total score of aligned group relatively large. We extend the vectorial inner product operation for more input and calculate the aligned score for each multimodal group. Considering that the computational complexity scales exponentially to the number of modalities, we adopt stochastic expectation approximation (SEA) for the real process. The extensive experimental results on several tasks reveal the effectiveness of our contributions.

A song is a combination of singing voice and accompaniment. However, existing works focus on singing voice synthesis and music generation independently. Little attention was paid to exploring song synthesis. In this work, we propose a novel task called Text-to-Song synthesis which incorporates both vocal and accompaniment generation. We develop Melodist, a two-stage text-to-song method that consists of singing voice synthesis (SVS) and vocal-to-accompaniment (V2A) synthesis. Melodist leverages tri-tower contrastive pretraining to learn more effective text representation for controllable V2A synthesis. A Chinese song dataset mined from a music website is built to alleviate data scarcity for our research. The evaluation results on our dataset demonstrate that Melodist can synthesize songs with comparable quality and style consistency. Audio samples can be found in https://text2songMelodist.github.io/Sample/.

Note-level Automatic Singing Voice Transcription (AST) converts singing recordings into note sequences, facilitating the automatic annotation of singing datasets for Singing Voice Synthesis (SVS) applications. Current AST methods, however, struggle with accuracy and robustness when used for practical annotation. This paper presents ROSVOT, the first robust AST model that serves SVS, incorporating a multi-scale framework that effectively captures coarse-grained note information and ensures fine-grained frame-level segmentation, coupled with an attention-based pitch decoder for reliable pitch prediction. We also established a comprehensive annotation-and-training pipeline for SVS to test the model in real-world settings. Experimental findings reveal that the proposed model achieves state-of-the-art transcription accuracy with either clean or noisy inputs. Moreover, when trained on enlarged, automatically annotated datasets, the SVS model outperforms its baseline, affirming the capability for practical application. Audio samples are available at https://rosvot.github.io. Codes can be found at https://github.com/RickyL-2000/ROSVOT.

In the field of speech synthesis, there is a growing emphasis on employing multimodal speech to enhance robustness. A key challenge in this area is the scarcity of datasets that pair audio with corresponding video. We employ a methodology that incorporates modality alignment during the pre-training phase on multimodal datasets, uniquely facilitating zero-shot generalization through the process of freezing the video modality feature extraction component and the encoder module within the pretrained weights, thereby enabling effective cross-modal and cross-lingual transfer. We have named this method ‘Uni-Dubbing’. Our method finely tunes with both multimodal and single-modality audio data. In multimodal scenarios, it achieves a reduced word error rate (WER) of 31.73%, surpassing the previous best of 33.9%. It also excels in metrics like tone quality and synchronization. With single-modality audio, it achieves a WER of 36.08%, demonstrating adaptability to limited data. Its domain generalization capabilities are proven across various language tasks in video translation and audio generation. Trained on 433 hours of audio data, it surpasses techniques using 200 hours of audiovisual data. The code and demo are available at https://diracer.github.io/unidubbing.

Large language models (LLMs) have successfully served as a general-purpose interface across multiple tasks and languages, while the adaptation of voice LLMs is mostly designed for specific purposes (either single-task or monolingual), where the advantages of LLMs especially for low-resource language processing and zero-shot task generalization are less exploited in the audio community. To bridge the gap, we introduce Make-A-Voice as a multi-modal voice LLM and conduct a comprehensive study on its capability to deal with multiple tasks/languages. When trained on ~200K hours of 6-language data for 4 voice generation applications, Make-A-Voice emerges notable advantages: 1) as scalable learners to improve performance with end-to-end local and global multiscale transformers; and 2) as multitask learners by adjusting prompts to share common knowledge across modalities (speech/singing) and present in-context learning abilities by generalizing to unseen tasks not explicitly train on; 3) as multilingual learners to alleviate data scarcity of low-resource languages by including rich-resource language training data. Experimental results demonstrate that Make-A-Voice exhibits superior audio quality and style similarity compared with competitive baseline models in monolingual/cross-lingual voice generation. Audio samples are available at https://M-Voice.github.io

Zero-shot text-to-speech (TTS) has gained significant attention due to its powerful voice cloning capabilities, requiring only a few seconds of unseen speaker voice prompts. However, all previous work has been developed for cloud-based systems. Taking autoregressive models as an example, although these approaches achieve high-fidelity voice cloning, they fall short in terms of inference speed, model size, and robustness. Therefore, we propose MobileSpeech, which is a fast, lightweight, and robust zero-shot text-to-speech system based on mobile devices for the first time. Specifically: 1) leveraging discrete codec, we design a parallel speech mask decoder module called SMD, which incorporates hierarchical information from the speech codec and weight mechanisms across different codec layers during the generation process. Moreover, to bridge the gap between text and speech, we introduce a high-level probabilistic mask that simulates the progression of information flow from less to more during speech generation. 2) For speaker prompts, we extract fine-grained prompt duration from the prompt speech and incorporate text, prompt speech by cross attention in SMD. We demonstrate the effectiveness of MobileSpeech on multilingual datasets at different levels, achieving state-of-the-art results in terms of generating speed and speech quality. MobileSpeech achieves RTF of 0.09 on a single A100 GPU and we have successfully deployed MobileSpeech on mobile devices. Audio samples are available at https://mobilespeech.github.io/

Direct speech-to-speech translation (S2ST) with discrete self-supervised representations has achieved remarkable accuracy, but is unable to preserve the speaker timbre of the source speech. Meanwhile, the scarcity of high-quality speaker-parallel data poses a challenge for learning style transfer during translation. We design an S2ST pipeline with style-transfer capability on the basis of discrete self-supervised speech representations and codec units. The acoustic language model we introduce for style transfer leverages self-supervised in-context learning, acquiring style transfer ability without relying on any speaker-parallel data, thereby overcoming data scarcity. By using extensive training data, our model achieves zero-shot cross-lingual style transfer on previously unseen source languages. Experiments show that our model generates translated speeches with high fidelity and speaker similarity. Audio samples are available at http://stylelm.github.io/ .

Argument mining aims to detect all possible argumentative components and identify their relationships automatically. As a thriving task in natural language processing, there has been a large amount of corpus for academic study and application development in this field. However, the research in this area is still constrained by the inherent limitations of existing datasets. Specifically, all the publicly available datasets are relatively small in scale, and few of them provide information from other modalities to facilitate the learning process. Moreover, the statements and expressions in these corpora are usually in a compact form, which restricts the generalization ability of models. To this end, we collect a novel dataset AntCritic to serve as a helpful complement to this area, which consists of about 10k free-form and visually-rich financial comments and supports both argument component detection and argument relation prediction tasks. Besides, to cope with the challenges brought by scenario expansion, we thoroughly explore the fine-grained relation prediction and structure reconstruction scheme and discuss the encoding mechanism for visual styles and layouts. On this basis, we design two simple but effective model architectures and conduct various experiments on this dataset to provide benchmark performances as a reference and verify the practicability of our proposed architecture. We release our data and code in this link, and this dataset follows CC BY-NC-ND 4.0 license.

Abductive Reasoning, has long been considered to be at the core ability of humans, which enables us to infer the most plausible explanation of incomplete known phenomena in daily life. However, such critical reasoning capability is rarely investigated for contemporary AI systems under such limited observations. To facilitate this research community, this paper sheds new light on Abductive Reasoning by studying a new vision-language task, Multi-modal Action chain abductive Reasoning (MAR), together with a large-scale Abductive Reasoning dataset: Given an incomplete set of language described events, MAR aims to imagine the most plausible event by spatio-temporal grounding in past video and then infer the hypothesis of subsequent action chain that can best explain the language premise. To solve this task, we propose a strong baseline model that realizes MAR from two perspectives: (i) we first introduce the transformer, which learns to encode the observation to imagine the plausible event with explicitly interpretable event grounding in the video based on the commonsense knowledge recognition ability. (ii) To complete the assumption of a follow-up action chain, we design a novel symbolic module that can complete strict derivation of the progressive action chain layer by layer. We conducted extensive experiments on the proposed dataset, and the experimental study shows that the proposed model significantly outperforms existing video-language models in terms of effectiveness on our newly created MAR dataset.

Speech Recognition builds a bridge between the multimedia streaming (audio-only, visual-only or audio-visual) and the corresponding text transcription. However, when training the specific model of new domain, it often gets stuck in the lack of new-domain utterances, especially the labeled visual utterances. To break through this restriction, we attempt to achieve zero-shot modality transfer by maintaining the multi-modality alignment in phoneme space learned with unlabeled multimedia utterances in the high resource domain during the pre-training, and propose a training system Open-modality Speech Recognition (OpenSR) that enables the models trained on a single modality (e.g., audio-only) applicable to more modalities (e.g., visual-only and audio-visual). Furthermore, we employ a cluster-based prompt tuning strategy to handle the domain shift for the scenarios with only common words in the new domain utterances. We demonstrate that OpenSR enables modality transfer from one to any in three different settings (zero-, few- and full-shot), and achieves highly competitive zero-shot performance compared to the existing few-shot and full-shot lip-reading methods. To the best of our knowledge, OpenSR achieves the state-of-the-art performance of word error rate in LRS2 on audio-visual speech recognition and lip-reading with 2.7% and 25.0%, respectively.

Direct speech-to-speech translation (S2ST) aims to convert speech from one language into another, and has demonstrated significant progress to date. Despite the recent success, current S2ST models still suffer from distinct degradation in noisy environments and fail to translate visual speech (i.e., the movement of lips and teeth). In this work, we present AV-TranSpeech, the first audio-visual speech-to-speech (AV-S2ST) translation model without relying on intermediate text. AV-TranSpeech complements the audio stream with visual information to promote system robustness and opens up a host of practical applications: dictation or dubbing archival films. To mitigate the data scarcity with limited parallel AV-S2ST data, we 1) explore self-supervised pre-training with unlabeled audio-visual data to learn contextual representation, and 2) introduce cross-modal distillation with S2ST models trained on the audio-only corpus to further reduce the requirements of visual data. Experimental results on two language pairs demonstrate that AV-TranSpeech outperforms audio-only models under all settings regardless of the type of noise. With low-resource audio-visual data (10h, 30h), cross-modal distillation yields an improvement of 7.6 BLEU on average compared with baselines. Audio samples are available at https://AV-TranSpeech.github.io/.

Improving text representation has attracted much attention to achieve expressive text-to-speech (TTS). However, existing works only implicitly learn the prosody with masked token reconstruction tasks, which leads to low training efficiency and difficulty in prosody modeling. We propose CLAPSpeech, a cross-modal contrastive pre-training framework that learns from the prosody variance of the same text token under different contexts. Specifically, 1) with the design of a text encoder and a prosody encoder, we encourage the model to connect the text context with its corresponding prosody pattern in the joint multi-modal space; 2) we introduce a multi-scale pre-training pipeline to capture prosody patterns in multiple levels. 3) we show how to incorporate CLAPSpeech into existing TTS models for better prosody. Experiments on three datasets not only show that CLAPSpeech could improve the prosody prediction for existing TTS methods, but also demonstrate its generalization ability to adapt to multiple languages and multi-speaker text-to-speech. We also deeply analyze the principle behind the performance of CLAPSpeech. Ablation studies demonstrate the necessity of each component in CLAPSpeech. Source code and audio samples are available at https://clapspeech.github.io.

The task of spoken video grounding aims to localize moments in videos that are relevant to descriptive spoken queries. However, extracting semantic information from speech and modeling the cross-modal correlation pose two critical challenges. Previous studies solve them by representing spoken queries based on the matched video frames, which require tremendous effort for frame-level labeling. In this work, we investigate weakly-supervised spoken video grounding, i.e., learning to localize moments without expensive temporal annotations. To effectively represent the cross-modal semantics, we propose Semantic Interaction Learning (SIL), a novel framework consisting of the acoustic-semantic pre-training (ASP) and acoustic-visual contrastive learning (AVCL). In ASP, we pre-train an effective encoder for the grounding task with three comprehensive tasks, where the robustness task enhances stability by explicitly capturing the invariance between time- and frequency-domain features, the conciseness task avoids over-smooth attention by compressing long sequence into segments, and the semantic task improves spoken language understanding by modeling the precise semantics. In AVCL, we mine pseudo labels with discriminative sampling strategies and directly strengthen the interaction between speech and video by maximizing their mutual information. Extensive experiments demonstrate the effectiveness and superiority of our method.

Audio-visual text generation aims to understand multi-modality contents and translate them into texts. Although various transfer learning techniques of text generation have been proposed, they focused on uni-modal analysis (e.g. text-to-text, visual-to-text) and lack consideration of multi-modal content and cross-modal relation. Motivated by the fact that humans can recognize the timbre of the same low-level concepts (e.g., footstep, rainfall, and laughing), even in different visual conditions, we aim to mitigate the domain discrepancies by audio-visual correlation. In this paper, we propose a novel Transferable Audio-Visual Text Generation framework, named TAVT, which consists of two key components: Audio-Visual Meta-Mapper (AVMM) and Dual Counterfactual Contrastive Learning (DCCL). (1) AVMM first introduces a universal auditory semantic space and drifts the domain-invariant low-level concepts into visual prefixes. Then the reconstruct-based learning encourages the AVMM to learn “which pixels belong to the same sound” and achieve audio-enhanced visual prefix. The well-trained AVMM can be further applied to uni-modal setting. (2) Furthermore, DCCL leverages the destructive counterfactual transformations to provide cross-modal constraints for AVMM from the perspective of feature distribution and text generation. (3) The experimental results show that TAVT outperforms the state-of-the-art methods across multiple domains (cross-datasets, cross-categories) and various modal settings (uni-modal, multi-modal).

Natural language video localization(NLVL) task involves the semantic matching of a text query with a moment from an untrimmed video. Previous methods primarily focus on improving performance with the assumption of independently identical data distribution while ignoring the out-of-distribution data. Therefore, these approaches often fail when handling the videos and queries in novel scenes, which is inevitable in real-world scenarios. In this paper, we, for the first time, formulate the scene-robust NLVL problem and propose a novel generalizable NLVL framework utilizing data in multiple available scenes to learn a robust model. Specifically, our model learns a group of generalizable domain-invariant representations by alignment and decomposition. First, we propose a comprehensive intra- and inter-sample distance metric for complex multi-modal feature space, and an asymmetric multi-modal alignment loss for different information densities of text and vision. Further, to alleviate the conflict between domain-invariant features for generalization and domain-specific information for reasoning, we introduce domain-specific and domain-agnostic predictors to decompose and refine the learned features by dynamically adjusting the weights of samples. Based on the original video tags, we conduct extensive experiments on three NLVL datasets with different-grained scene shifts to show the effectiveness of our proposed methods.

We are interested in a challenging task, Realistic-Music-Score based Singing Voice Synthesis (RMS-SVS). RMS-SVS aims to generate high-quality singing voices given realistic music scores with different note types (grace, slur, rest, etc.). Though significant progress has been achieved, recent singing voice synthesis (SVS) methods are limited to fine-grained music scores, which require a complicated data collection pipeline with time-consuming manual annotation to align music notes with phonemes. % Furthermore, existing approaches cannot synthesize rhythmic singing voices given realistic music scores due to the domain gap between fine-grained music scores and realistic music scores. Furthermore, these manual annotation destroys the regularity of note durations in music scores, making fine-grained music scores inconvenient for composing. To tackle these challenges, we propose RMSSinger, the first RMS-SVS method, which takes realistic music scores as input, eliminating most of the tedious manual annotation and avoiding the aforementioned inconvenience. Note that music scores are based on words rather than phonemes, in RMSSinger, we introduce word-level modeling to avoid the time-consuming phoneme duration annotation and the complicated phoneme-level mel-note alignment. Furthermore, we propose the first diffusion-based pitch modeling method, which ameliorates the naturalness of existing pitch-modeling methods. To achieve these, we collect a new dataset containing realistic music scores and singing voices according to these realistic music scores from professional singers. Extensive experiments on the dataset demonstrate the effectiveness of our methods. Audio samples are available at https://rmssinger.github.io/.

Generative adversarial networks (GANs) and denoising diffusion probabilistic models (DDPMs) have recently achieved impressive performances in image and audio synthesis. After revisiting their success in conditional speech synthesis, we find that 1) GANs sacrifice sample diversity for quality and speed, 2) diffusion models exhibit outperformed sample quality and diversity at a high computational cost, where achieving high-quality, fast, and diverse speech synthesis challenges all neural synthesizers. In this work, we propose to converge advantages from GANs and diffusion models by incorporating both classes, introducing dual-empowered modeling perspectives: 1) FastDiff 2 (DiffGAN), a diffusion model whose denoising process is parametrized by conditional GANs, and the non-Gaussian denoising distribution makes it much more stable to implement the reverse process with large steps sizes; and 2) FastDiff 2 (GANDiff), a generative adversarial network whose forward process is constructed by multiple denoising diffusion iterations, which exhibits better sample diversity than traditional GANs. Experimental results show that both variants enjoy an efficient 4-step sampling process and demonstrate superior sample quality and diversity. Audio samples are available at https://RevisitSpeech.github.io/

The speech-to-singing (STS) voice conversion task aims to generate singing samples corresponding to speech recordings while facing a major challenge: the alignment between the target (singing) pitch contour and the source (speech) content is difficult to learn in a text-free situation. This paper proposes AlignSTS, an STS model based on explicit cross-modal alignment, which views speech variance such as pitch and content as different modalities. Inspired by the mechanism of how humans will sing the lyrics to the melody, AlignSTS: 1) adopts a novel rhythm adaptor to predict the target rhythm representation to bridge the modality gap between content and pitch, where the rhythm representation is computed in a simple yet effective way and is quantized into a discrete space; and 2) uses the predicted rhythm representation to re-align the content based on cross-attention and conducts a cross-modal fusion for re-synthesize. Extensive experiments show that AlignSTS achieves superior performance in terms of both objective and subjective metrics. Audio samples are available at https://alignsts.github.io.

Expressive text-to-speech aims to generate high-quality samples with rich and diverse prosody, which is hampered by dual challenges: 1) prosodic attributes in highly dynamic voices are difficult to capture and model without intonation; and 2) highly multimodal prosodic representations cannot be well learned by simple regression (e.g., MSE) objectives, which causes blurry and over-smoothing predictions. This paper proposes Prosody-TTS, a two-stage pipeline that enhances prosody modeling and sampling by introducing several components: 1) a self-supervised masked autoencoder to model the prosodic representation without relying on text transcriptions or local prosody attributes, which ensures to cover diverse speaking voices with superior generalization; and 2) a diffusion model to sample diverse prosodic patterns within the latent space, which prevents TTS models from generating samples with dull prosodic performance. Experimental results show that Prosody-TTS achieves new state-of-the-art in text-to-speech with natural and expressive synthesis. Both subjective and objective evaluation demonstrate that it exhibits superior audio quality and prosody naturalness with rich and diverse prosodic attributes. Audio samples are available at https://improved_prosody.github.io

Visual segmentation from language queries has attracted significant research interest. Despite the effectiveness, existing works require expensive labeling and suffer severe degradation when deployed to an unseen domain. In this paper, we investigate a novel task Cross-domain Query-based Visual Segmentation (CQVS), aiming to adapt the segmentation model from a labeled domain to a new unlabeled domain. The challenges of CQVS stem from three domain discrepancies: (1) multi-modal content shift, (2) uni-modal feature gap and (3) cross-modal relation bias. Existing domain adaptation methods fail to address them comprehensively and precisely (e.g. at pixel level), thus being suboptimal for CQVS. To overcome this limitation, we propose Semantic-conditioned Dual Adaptation (SDA), a novel framework to achieve precise feature- and relation-invariant across domains via a universal semantic structure. The SDA consists of two key components: Content-aware Semantic Modeling (CSM) and Dual Adaptive Branches (DAB). First, CSM introduces a common semantic space across domains to provide uniform guidance. Then, DAB seamlessly leverages this semantic information to develop a contrastive feature branch for category-wise pixel alignment, and design a reciprocal relation branch for relation enhancement via two complementary masks. Extensive experiments on three video benchmarks and three image benchmarks evidence the superiority of our approach over the state-of-the-arts.

Visual temporal-aligned translation aims to transform the visual sequence into natural words, including important applicable tasks such as lipreading and fingerspelling recognition. However, various performance habits of specific words by different speakers or signers can lead to visual ambiguity, which has become a major obstacle to the development of current methods. Considering the constraints above, the generalization ability of the translation system is supposed to be further explored through the evaluation results on unseen performers. In this paper, we develop a novel generalizable framework named Contrastive Token-Wise Meta-learning (CtoML), which strives to transfer recognition skills to unseen performers. To the best of our knowledge, employing meta-learning methods directly in the image domain poses two main challenges, and we propose corresponding strategies. First, sequence prediction in visual temporal-aligned translation, which aims to generate multiple words autoregressively, is different from the vanilla classification. Thus, we devise the token-wise diversity-aware weights for the meta-train stage, which encourages the model to make efforts on those ambiguously recognized tokens. Second, considering the consistency of word-visual prototypes across different domains, we develop two complementary global and local contrastive losses to maintain inter-class relationships and promote domain-independent. We conduct extensive experiments on the widely-used lipreading dataset GRID and the fingerspelling dataset ChicagoFSWild, and the experimental results show the effectiveness of our proposed CtoML over existing state-of-the-art methods.

Stutter removal is an essential scenario in the field of speech editing. However, when the speech recording contains stutters, the existing text-based speech editing approaches still suffer from: 1) the over-smoothing problem in the edited speech; 2) lack of robustness due to the noise introduced by stutter; 3) to remove the stutters, users are required to determine the edited region manually. To tackle the challenges in stutter removal, we propose FluentSpeech, a stutter-oriented automatic speech editing model. Specifically, 1) we propose a context-aware diffusion model that iteratively refines the modified mel-spectrogram with the guidance of context features; 2) we introduce a stutter predictor module to inject the stutter information into the hidden sequence; 3) we also propose a stutter-oriented automatic speech editing (SASE) dataset that contains spontaneous speech recordings with time-aligned stutter labels to train the automatic stutter localization model. Experimental results on VCTK and LibriTTS datasets demonstrate that our model achieves state-of-the-art performance on speech editing. Further experiments on our SASE dataset show that FluentSpeech can effectively improve the fluency of stuttering speech in terms of objective and subjective metrics. Code and audio samples can be found at https://github.com/Zain-Jiang/Speech-Editing-Toolkit.

Recently, binaural audio synthesis (BAS) has emerged as a promising research field for its applications in augmented and virtual realities. Binaural audio helps ususers orient themselves and establish immersion by providing the brain with interaural time differences reflecting spatial information. However, existing BAS methods are limited in terms of phase estimation, which is crucial for spatial hearing. In this paper, we propose the DopplerBAS method to explicitly address the Doppler effect of the moving sound source. Specifically, we calculate the radial relative velocity of the moving speaker in spherical coordinates, which further guides the synthesis of binaural audio. This simple method introduces no additional hyper-parameters and does not modify the loss functions, and is plug-and-play: it scales well to different types of backbones. DopperBAS distinctly improves the representative WarpNet and BinauralGrad backbones in the phase error metric and reaches a new state of the art (SOTA): 0.780 (versus the current SOTA 0.807). Experiments and ablation studies demonstrate the effectiveness of our method.

Deductive reasoning is a crucial cognitive ability of humanity, allowing us to derive valid conclusions from premises and observations. However, existing works mainly focus on language-based premises and generally neglect deductive reasoning from visual observations. In this work, we introduce rule bAsed futuRe-inference deducTion (ART), which aims at deducing the correct future event based on the visual phenomenon (a video) and the rule-based premises, along with an explanation of the reasoning process. To advance this field, we construct a large-scale densely annotated dataset (Video-ART), where the premises, future event candidates, the reasoning process explanation, and auxiliary commonsense knowledge (e.g., actions and appearance) are annotated by annotators. Upon Video-ART, we develop a strong baseline named ARTNet. In essence, guided by commonsense knowledge, ARTNet learns to identify the target video character and perceives its visual clues related to the future event. Then, ARTNet rigorously applies the given premises to conduct reasoning from the identified information to future events, through a non-parametric rule reasoning network and a reasoning-path review module. Empirical studies validate the rationality of ARTNet in deductive reasoning upon visual observations and the effectiveness over existing works.

3D visual grounding aims to localize the target object in a 3D point cloud by a free-form language description. Typically, the sentences describing the target object tend to provide information about its relative relation between other objects and its position within the whole scene. In this work, we propose a relation-aware one-stage framework, named 3D Relative Position-aware Network (3DRP-Net), which can effectively capture the relative spatial relationships between objects and enhance object attributes. Specifically, 1) we propose a 3D Relative Position Multi-head Attention (3DRP-MA) module to analyze relative relations from different directions in the context of object pairs, which helps the model to focus on the specific object relations mentioned in the sentence. 2) We designed a soft-labeling strategy to alleviate the spatial ambiguity caused by redundant points, which further stabilizes and enhances the learning process through a constant and discriminative distribution. Extensive experiments conducted on three benchmarks (i.e., ScanRefer and Nr3D/Sr3D) demonstrate that our method outperforms all the state-of-the-art methods in general.

Text-to-speech(TTS) has undergone remarkable improvements in performance, particularly with the advent of Denoising Diffusion Probabilistic Models (DDPMs). However, the perceived quality of audio depends not solely on its content, pitch, rhythm, and energy, but also on the physical environment.In this work, we propose ViT-TTS, the first visual TTS model with scalable diffusion transformers. ViT-TTS complement the phoneme sequence with the visual information to generate high-perceived audio, opening up new avenues for practical applications of AR and VR to allow a more immersive and realistic audio experience. To mitigate the data scarcity in learning visual acoustic information, we 1) introduce a self-supervised learning framework to enhance both the visual-text encoder and denoiser decoder; 2) leverage the diffusion transformer scalable in terms of parameters and capacity to learn visual scene information. Experimental results demonstrate that ViT-TTS achieves new state-of-the-art results, outperforming cascaded systems and other baselines regardless of the visibility of the scene. With low-resource data (1h, 2h, 5h), ViT-TTS achieves comparative results with rich-resource baselines.

We are interested in a novel task, singing voice beautification (SVB). Given the singing voice of an amateur singer, SVB aims to improve the intonation and vocal tone of the voice, while keeping the content and vocal timbre. Current automatic pitch correction techniques are immature, and most of them are restricted to intonation but ignore the overall aesthetic quality. Hence, we introduce Neural Singing Voice Beautifier (NSVB), the first generative model to solve the SVB task, which adopts a conditional variational autoencoder as the backbone and learns the latent representations of vocal tone. In NSVB, we propose a novel time-warping approach for pitch correction: Shape-Aware Dynamic Time Warping (SADTW), which ameliorates the robustness of existing time-warping approaches, to synchronize the amateur recording with the template pitch curve. Furthermore, we propose a latent-mapping algorithm in the latent space to convert the amateur vocal tone to the professional one. To achieve this, we also propose a new dataset containing parallel singing recordings of both amateur and professional versions. Extensive experiments on both Chinese and English songs demonstrate the effectiveness of our methods in terms of both objective and subjective metrics. Audio samples are available at https://neuralsvb.github.io. Codes: https://github.com/MoonInTheRiver/NeuralSVB.

Non-autoregressive text to speech (NAR-TTS) models have attracted much attention from both academia and industry due to their fast generation speed. One limitation of NAR-TTS models is that they ignore the correlation in time and frequency domains while generating speech mel-spectrograms, and thus cause blurry and over-smoothed results. In this work, we revisit this over-smoothing problem from a novel perspective: the degree of over-smoothness is determined by the gap between the complexity of data distributions and the capability of modeling methods. Both simplifying data distributions and improving modeling methods can alleviate the problem. Accordingly, we first study methods reducing the complexity of data distributions. Then we conduct a comprehensive study on NAR-TTS models that use some advanced modeling methods. Based on these studies, we find that 1) methods that provide additional condition inputs reduce the complexity of data distributions to model, thus alleviating the over-smoothing problem and achieving better voice quality. 2) Among advanced modeling methods, Laplacian mixture loss performs well at modeling multimodal distributions and enjoys its simplicity, while GAN and Glow achieve the best voice quality while suffering from increased training or model complexity. 3) The two categories of methods can be combined to further alleviate the over-smoothness and improve the voice quality. 4) Our experiments on the multi-speaker dataset lead to similar conclusions as above and providing more variance information can reduce the difficulty of modeling the target data distribution and alleviate the requirements for model capacity.

Natural language spatial video grounding aims to detect the relevant objects in video frames with descriptive sentences as the query. In spite of the great advances, most existing methods rely on dense video frame annotations, which require a tremendous amount of human effort. To achieve effective grounding under a limited annotation budget, we investigate one-shot video grounding and learn to ground natural language in all video frames with solely one frame labeled, in an end-to-end manner. One major challenge of end-to-end one-shot video grounding is the existence of videos frames that are either irrelevant to the language query or the labeled frame. Another challenge relates to the limited supervision, which might result in ineffective representation learning. To address these challenges, we designed an end-to-end model via Information Tree for One-Shot video grounding (IT-OS). Its key module, the information tree, can eliminate the interference of irrelevant frames based on branch search and branch cropping techniques. In addition, several self-supervised tasks are proposed based on the information tree to improve the representation learning under insufficient labeling. Experiments on the benchmark dataset demonstrate the effectiveness of our model.

This paper attacks the challenging problem of sign language translation (SLT), which involves not only visual and textual understanding but also additional prior knowledge learning (i.e. performing style, syntax). However, the majority of existing methods with vanilla encoder-decoder structures fail to sufficiently explore all of them. Based on this concern, we propose a novel method called Prior knowledge and memory Enriched Transformer (PET) for SLT, which incorporates the auxiliary information into vanilla transformer. Concretely, we develop gated interactive multi-head attention which associates the multimodal representation and global signing style with adaptive gated functions. One Part-of-Speech (POS) sequence generator relies on the associated information to predict the global syntactic structure, which is thereafter leveraged to guide the sentence generation. Besides, considering that the visual-textual context information, and additional auxiliary knowledge of a word may appear in more than one video, we design a multi-stream memory structure to obtain higher-quality translations, which stores the detailed correspondence between a word and its various relevant information, leading to a more comprehensive understanding for each word. We conduct extensive empirical studies on RWTH-PHOENIX-Weather-2014 dataset with both signer-dependent and signer-independent conditions. The quantitative and qualitative experimental results comprehensively reveal the effectiveness of PET.

Non-autoregressive (NAR) models generate all the tokens of a sequence in parallel, resulting in faster generation speed compared to their autoregressive (AR) counterparts but at the cost of lower accuracy. Different techniques including knowledge distillation and source-target alignment have been proposed to bridge the gap between AR and NAR models in various tasks such as neural machine translation (NMT), automatic speech recognition (ASR), and text to speech (TTS). With the help of those techniques, NAR models can catch up with the accuracy of AR models in some tasks but not in some others. In this work, we conduct a study to understand the difficulty of NAR sequence generation and try to answer: (1) Why NAR models can catch up with AR models in some tasks but not all? (2) Why techniques like knowledge distillation and source-target alignment can help NAR models. Since the main difference between AR and NAR models is that NAR models do not use dependency among target tokens while AR models do, intuitively the difficulty of NAR sequence generation heavily depends on the strongness of dependency among target tokens. To quantify such dependency, we propose an analysis model called CoMMA to characterize the difficulty of different NAR sequence generation tasks. We have several interesting findings: 1) Among the NMT, ASR and TTS tasks, ASR has the most target-token dependency while TTS has the least. 2) Knowledge distillation reduces the target-token dependency in target sequence and thus improves the accuracy of NAR models. 3) Source-target alignment constraint encourages dependency of a target token on source tokens and thus eases the training of NAR models.

In this work, we develop SimulSpeech, an end-to-end simultaneous speech to text translation system which translates speech in source language to text in target language concurrently. SimulSpeech consists of a speech encoder, a speech segmenter and a text decoder, where 1) the segmenter builds upon the encoder and leverages a connectionist temporal classification (CTC) loss to split the input streaming speech in real time, 2) the encoder-decoder attention adopts a wait-k strategy for simultaneous translation. SimulSpeech is more challenging than previous cascaded systems (with simultaneous automatic speech recognition (ASR) and simultaneous neural machine translation (NMT)). We introduce two novel knowledge distillation methods to ensure the performance: 1) Attention-level knowledge distillation transfers the knowledge from the multiplication of the attention matrices of simultaneous NMT and ASR models to help the training of the attention mechanism in SimulSpeech; 2) Data-level knowledge distillation transfers the knowledge from the full-sentence NMT model and also reduces the complexity of data distribution to help on the optimization of SimulSpeech. Experiments on MuST-C English-Spanish and English-German spoken language translation datasets show that SimulSpeech achieves reasonable BLEU scores and lower delay compared to full-sentence end-to-end speech to text translation (without simultaneous translation), and better performance than the two-stage cascaded simultaneous translation model in terms of BLEU scores and translation delay.

Video dialog is a new and challenging task, which requires the agent to answer questions combining video information with dialog history. And different from single-turn video question answering, the additional dialog history is important for video dialog, which often includes contextual information for the question. Existing visual dialog methods mainly use RNN to encode the dialog history as a single vector representation, which might be rough and straightforward. Some more advanced methods utilize hierarchical structure, attention and memory mechanisms, which still lack an explicit reasoning process. In this paper, we introduce a novel progressive inference mechanism for video dialog, which progressively updates query information based on dialog history and video content until the agent think the information is sufficient and unambiguous. In order to tackle the multi-modal fusion problem, we propose a cross-transformer module, which could learn more fine-grained and comprehensive interactions both inside and between the modalities. And besides answer generation, we also consider question generation, which is more challenging but significant for a complete video dialog system. We evaluate our method on two large-scale datasets, and the extensive experiments show the effectiveness of our method.

Natural Language Inference (NLI), also known as Recognizing Textual Entailment (RTE), is one of the most important problems in natural language processing. It requires to infer the logical relationship between two given sentences. While current approaches mostly focus on the interaction architectures of the sentences, in this paper, we propose to transfer knowledge from some important discourse markers to augment the quality of the NLI model. We observe that people usually use some discourse markers such as “so” or “but” to represent the logical relationship between two sentences. These words potentially have deep connections with the meanings of the sentences, thus can be utilized to help improve the representations of them. Moreover, we use reinforcement learning to optimize a new objective function with a reward defined by the property of the NLI datasets to make full use of the labels information. Experiments show that our method achieves the state-of-the-art performance on several large-scale datasets.

In this study, we explore capsule networks with dynamic routing for text classification. We propose three strategies to stabilize the dynamic routing process to alleviate the disturbance of some noise capsules which may contain “background” information or have not been successfully trained. A series of experiments are conducted with capsule networks on six text classification benchmarks. Capsule networks achieve state of the art on 4 out of 6 datasets, which shows the effectiveness of capsule networks for text classification. We additionally show that capsule networks exhibit significant improvement when transfer single-label to multi-label text classification over strong baseline methods. To the best of our knowledge, this is the first work that capsule networks have been empirically investigated for text modeling.

We study the problem of identifying the topics and sentiments and tracking their shifts from social media texts in different geographical regions during emergencies and disasters. We propose a location-based dynamic sentiment-topic model (LDST) which can jointly model topic, sentiment, time and Geolocation information. The experimental results demonstrate that LDST performs very well at discovering topics and sentiments from social media and tracking their shifts in different geographical regions during emergencies and disasters. We will release the data and source code after this work is published.