Li-Wei Chen

Li-Wei Chen, Shinji Watanabe, Alexander Rudnicky

Recent Text-to-Speech (TTS) systems trained on reading or acted corpora have achieved near human-level naturalness. The diversity of human speech, however, often goes beyond the coverage of these corpora. We believe the ability to handle such diversity is crucial for AI systems to achieve human-level communication. Our work explores the use of more abundant real-world data for building speech synthesizers. We train TTS systems using real-world speech from YouTube and podcasts. We observe the mismatch between training and inference alignments in mel-spectrogram based autoregressive models, leading to unintelligible synthesis, and demonstrate that learned discrete codes within multiple code groups effectively resolves this issue. We introduce our MQTTS system whose architecture is designed for multiple code generation and monotonic alignment, along with the use of a clean silence prompt to improve synthesis quality. We conduct ablation analyses to identify the efficacy of our methods. We show that MQTTS outperforms existing TTS systems in several objective and subjective measures.

Li-Wei Chen, Shinji Watanabe, Alexander Rudnicky

We present a unified system to realize one-shot voice conversion (VC) on the pitch, rhythm, and speaker attributes. Existing works generally ignore the correlation between prosody and language content, leading to the degradation of naturalness in converted speech. Additionally, the lack of proper language features prevents these systems from accurately preserving language content after conversion. To address these issues, we devise a cascaded modular system leveraging self-supervised discrete speech units as language representation. These discrete units provide duration information essential for rhythm modeling. Our system first extracts utterance-level prosody and speaker representations from the raw waveform. Given the prosody representation, a prosody predictor estimates pitch, energy, and duration for each discrete unit in the utterance. A synthesizer further reconstructs speech based on the predicted prosody, speaker representation, and discrete units. Experiments show that our system outperforms previous approaches in naturalness, intelligibility, speaker transferability, and prosody transferability. Code and samples are publicly available.

Li-Wei Chen, Yao-Fei Cheng, Hung-Shin Lee et al.

The lack of clean speech is a practical challenge to the development of speech enhancement systems, which means that there is an inevitable mismatch between their training criterion and evaluation metric. In response to this unfavorable situation, we propose a training and inference strategy that additionally uses enhanced speech as a target by improving the previously proposed noisy-target training (NyTT). Because homogeneity between in-domain noise and extraneous noise is the key to the effectiveness of NyTT, we train various student models by remixing 1) the teacher model's estimated speech and noise for enhanced-target training or 2) raw noisy speech and the teacher model's estimated noise for noisy-target training. Experimental results show that our proposed method outperforms several baselines, especially with the teacher/student inference, where predicted clean speech is derived successively through the teacher and final student models.

Georg Kohl, Li-Wei Chen, Nils Thuerey

Simulating turbulent flows is crucial for a wide range of applications, and machine learning-based solvers are gaining increasing relevance. However, achieving temporal stability when generalizing to longer rollout horizons remains a persistent challenge for learned PDE solvers. In this work, we analyze if fully data-driven fluid solvers that utilize an autoregressive rollout based on conditional diffusion models are a viable option to address this challenge. We investigate accuracy, posterior sampling, spectral behavior, and temporal stability, while requiring that methods generalize to flow parameters beyond the training regime. To quantitatively and qualitatively benchmark the performance of various flow prediction approaches, three challenging 2D scenarios including incompressible and transonic flows, as well as isotropic turbulence are employed. We find that even simple diffusion-based approaches can outperform multiple established flow prediction methods in terms of accuracy and temporal stability, while being on par with state-of-the-art stabilization techniques like unrolling at training time. Such traditional architectures are superior in terms of inference speed, however, the probabilistic nature of diffusion approaches allows for inferring multiple predictions that align with the statistics of the underlying physics. Overall, our benchmark contains three carefully chosen data sets that are suitable for probabilistic evaluation alongside various established flow prediction architectures.

Ombretta Strafforello, Gonzalo Muradas Odriozola, Fatemeh Behrad et al.

Assessing the aesthetic quality of artistic images presents unique challenges due to the subjective nature of aesthetics and the complex visual characteristics inherent to artworks. Basic data augmentation techniques commonly applied to natural images in computer vision may not be suitable for art images in aesthetic evaluation tasks, as they can change the composition of the art images. In this paper, we explore the impact of local and global data augmentation techniques on artistic image aesthetic assessment (IAA). We introduce BackFlip, a local data augmentation technique designed specifically for artistic IAA. We evaluate the performance of BackFlip across three artistic image datasets and four neural network architectures, comparing it with the commonly used data augmentation techniques. Then, we analyze the effects of components within the BackFlip pipeline through an ablation study. Our findings demonstrate that local augmentations, such as BackFlip, tend to outperform global augmentations on artistic IAA in most cases, probably because they do not perturb the composition of the art images. These results emphasize the importance of considering both local and global augmentations in future computational aesthetics research.

Anne-Sofie Maerten, Li-Wei Chen, Stefanie De Winter et al.

We present the Leuven Art Personalized Image Set (LAPIS), a novel dataset for personalized image aesthetic assessment (PIAA). It is the first dataset with images of artworks that is suitable for PIAA. LAPIS consists of 11,723 images and was meticulously curated in collaboration with art historians. Each image has an aesthetics score and a set of image attributes known to relate to aesthetic appreciation. Besides rich image attributes, LAPIS offers rich personal attributes of each annotator. We implemented two existing state-of-the-art PIAA models and assessed their performance on LAPIS. We assess the contribution of personal attributes and image attributes through ablation studies and find that performance deteriorates when certain personal and image attributes are removed. An analysis of failure cases reveals that both existing models make similar incorrect predictions, highlighting the need for improvements in artistic image aesthetic assessment. The LAPIS project page can be found at: https://github.com/Anne-SofieMaerten/LAPIS

Li-Wei Chen, Takuya Higuchi, Zakaria Aldeneh et al.

The success of large language models in text processing has inspired their adaptation to speech modeling. However, since speech is continuous and complex, it is often discretized for autoregressive modeling. Speech tokens derived from self-supervised models (known as semantic tokens) typically focus on the linguistic aspects of speech but neglect prosodic information. As a result, models trained on these tokens can generate speech with reduced naturalness. Existing approaches try to fix this by adding pitch features to the semantic tokens. However, pitch alone cannot fully represent the range of paralinguistic attributes, and selecting the right features requires careful hand-engineering. To overcome this, we propose an end-to-end variational approach that automatically learns to encode these continuous speech attributes to enhance the semantic tokens. Our approach eliminates the need for manual extraction and selection of paralinguistic features. Moreover, it produces preferred speech continuations according to human raters. Code, samples and models are available at https://github.com/b04901014/vae-gslm.

Han-Yu Lin, Li-Wei Chen, Hung-Shin Lee

Compression artifacts from standard video codecs often degrade perceptual quality. We propose a lightweight, semantic-aware pre-processing framework that enhances perceptual fidelity by selectively addressing these distortions. Our method integrates semantic embeddings from a vision-language model into an efficient convolutional architecture, prioritizing the preservation of perceptually significant structures. The model is trained end-to-end with a differentiable codec proxy, enabling it to mitigate artifacts from various standard codecs without modifying the existing video pipeline. During inference, the codec proxy is discarded, and SCENE operates as a standalone pre-processor, enabling real-time performance. Experiments on high-resolution benchmarks show improved performance over baselines in both objective (MS-SSIM) and perceptual (VMAF) metrics, with notable gains in preserving detailed textures within salient regions. Our results show that semantic-guided, codec-aware pre-processing is an effective approach for enhancing compressed video streams.

Chien-Chun Wang, Li-Wei Chen, Hung-Shin Lee et al.

Cross-domain speech enhancement (SE) is often faced with severe challenges due to the scarcity of noise and background information in an unseen target domain, leading to a mismatch between training and test conditions. This study puts forward a novel data simulation method to address this issue, leveraging noise-extractive techniques and generative adversarial networks (GANs) with only limited target noisy speech data. Notably, our method employs a noise encoder to extract noise embeddings from target-domain data. These embeddings aptly guide the generator to synthesize utterances acoustically fitted to the target domain while authentically preserving the phonetic content of the input clean speech. Furthermore, we introduce the notion of dynamic stochastic perturbation, which can inject controlled perturbations into the noise embeddings during inference, thereby enabling the model to generalize well to unseen noise conditions. Experiments on the VoiceBank-DEMAND benchmark dataset demonstrate that our domain-adaptive SE method outperforms an existing strong baseline based on data simulation.

Li-Wei Chen, Ombretta Strafforello, Anne-Sofie Maerten et al.

Image aesthetic assessment (IAA) evaluates image aesthetics, a task complicated by image diversity and user subjectivity. Current approaches address this in two stages: Generic IAA (GIAA) models estimate mean aesthetic scores, while Personal IAA (PIAA) models adapt GIAA using transfer learning to incorporate user subjectivity. However, a theoretical understanding of transfer learning between GIAA and PIAA, particularly concerning the impact of group composition, group size, aesthetic differences between groups and individuals, and demographic correlations, is lacking. This work establishes a theoretical foundation for IAA, proposing a unified model that encodes individual characteristics in a distributional format for both individual and group assessments. We show that transferring from GIAA to PIAA involves extrapolation, while the reverse involves interpolation, which is generally more effective for machine learning. Extensive experiments with varying group compositions, including sub-sampling by group size and disjoint demographics, reveal substantial performance variation even for GIAA, challenging the assumption that averaging scores eliminates individual subjectivity. Score-distribution analysis using Earth Mover's Distance (EMD) and the Gini index identifies education, photography experience, and art experience as key factors in aesthetic differences, with greater subjectivity in artworks than in photographs. Code is available at https://github.com/lwchen6309/aesthetics_transfer_learning.

Chien-Chun Wang, Li-Wei Chen, Cheng-Kang Chou et al.

While pre-trained automatic speech recognition (ASR) systems demonstrate impressive performance on matched domains, their performance often degrades when confronted with channel mismatch stemming from unseen recording environments and conditions. To mitigate this issue, we propose a novel channel-aware data simulation method for robust ASR training. Our method harnesses the synergistic power of channel-extractive techniques and generative adversarial networks (GANs). We first train a channel encoder capable of extracting embeddings from arbitrary audio. On top of this, channel embeddings are extracted using a minimal amount of target-domain data and used to guide a GAN-based speech synthesizer. This synthesizer generates speech that faithfully preserves the phonetic content of the input while mimicking the channel characteristics of the target domain. We evaluate our method on the challenging Hakka Across Taiwan (HAT) and Taiwanese Across Taiwan (TAT) corpora, achieving relative character error rate (CER) reductions of 20.02% and 9.64%, respectively, compared to the baselines. These results highlight the efficacy of our channel-aware data simulation method for bridging the gap between source- and target-domain acoustics.

Yao-Fei Cheng, Li-Wei Chen, Hung-Shin Lee et al.

This study investigates the efficacy of data augmentation techniques for low-resource automatic speech recognition (ASR), focusing on two endangered Austronesian languages, Amis and Seediq. Recognizing the potential of self-supervised learning (SSL) in low-resource settings, we explore the impact of data volume on the continued pre-training of SSL models. We propose a novel data-selection scheme leveraging a multilingual corpus to augment the limited target language data. This scheme utilizes a language classifier to extract utterance embeddings and employs one-class classifiers to identify utterances phonetically and phonologically proximate to the target languages. Utterances are ranked and selected based on their decision scores, ensuring the inclusion of highly relevant data in the SSL-ASR pipeline. Our experimental results demonstrate the effectiveness of this approach, yielding substantial improvements in ASR performance for both Amis and Seediq. These findings underscore the feasibility and promise of data augmentation through cross-lingual transfer learning for low-resource language ASR.

Cheng-Yeh Yang, Chien-Chun Wang, Li-Wei Chen et al.

Low-resource automatic speech recognition (ASR) continues to pose significant challenges, primarily due to the limited availability of transcribed data for numerous languages. While a wealth of spoken content is accessible in television dramas and online videos, Taiwanese Hokkien exemplifies this issue, with transcriptions often being scarce and the majority of available subtitles provided only in Mandarin. To address this deficiency, we introduce TG-ASR for Taiwanese Hokkien drama speech recognition, a translation-guided ASR framework that utilizes multilingual translation embeddings to enhance recognition performance in low-resource environments. The framework is centered around the parallel gated cross-attention (PGCA) mechanism, which adaptively integrates embeddings from various auxiliary languages into the ASR decoder. This mechanism facilitates robust cross-linguistic semantic guidance while ensuring stable optimization and minimizing interference between languages. To support ongoing research initiatives, we present YT-THDC, a 30-hour corpus of Taiwanese Hokkien drama speech with aligned Mandarin subtitles and manually verified Taiwanese Hokkien transcriptions. Comprehensive experiments and analyses identify the auxiliary languages that most effectively enhance ASR performance, achieving a 14.77% relative reduction in character error rate and demonstrating the efficacy of translation-guided learning for underrepresented languages in practical applications.

Pei-Jun Liao, Hung-Shin Lee, Yao-Fei Cheng et al.

Predicting stock prices presents challenges in financial forecasting. While traditional approaches such as ARIMA and RNNs are prevalent, recent developments in Large Language Models (LLMs) offer alternative methodologies. This paper introduces an approach that integrates LLMs with daily financial news for stock price prediction. To address the challenge of processing news data and identifying relevant content, we utilize stock name embeddings within attention mechanisms. Specifically, we encode news articles using a pre-trained LLM and implement three attention-based pooling techniques -- self-attentive, cross-attentive, and position-aware self-attentive pooling -- to filter news based on stock relevance. The filtered news embeddings, combined with historical stock prices, serve as inputs to the prediction model. Unlike prior studies that focus on individual stocks, our method trains a single generalized model applicable across multiple stocks. Experimental results demonstrate a 7.11% reduction in Mean Absolute Error (MAE) compared to the baseline, indicating the utility of stock name embeddings for news filtering and price forecasting within a generalized framework.

Li-Wei Chen, Kai-Chen Cheng, Hung-Shin Lee

This report provides a concise overview of the proposed North system, which aims to achieve automatic word/syllable recognition for Taiwanese Hakka (Sixian). The report outlines three key components of the system: the acquisition, composition, and utilization of the training data; the architecture of the model; and the hardware specifications and operational statistics. The demonstration of the system has been made public at https://asrvm.iis.sinica.edu.tw/hakka_sixian.

Li-Wei Chen, Hung-Shin Lee, Chen-Chi Chang

This paper introduces VoxHakka, a text-to-speech (TTS) system designed for Taiwanese Hakka, a critically under-resourced language spoken in Taiwan. Leveraging the YourTTS framework, VoxHakka achieves high naturalness and accuracy and low real-time factor in speech synthesis while supporting six distinct Hakka dialects. This is achieved by training the model with dialect-specific data, allowing for the generation of speaker-aware Hakka speech. To address the scarcity of publicly available Hakka speech corpora, we employed a cost-effective approach utilizing a web scraping pipeline coupled with automatic speech recognition (ASR)-based data cleaning techniques. This process ensured the acquisition of a high-quality, multi-speaker, multi-dialect dataset suitable for TTS training. Subjective listening tests conducted using comparative mean opinion scores (CMOS) demonstrate that VoxHakka significantly outperforms existing publicly available Hakka TTS systems in terms of pronunciation accuracy, tone correctness, and overall naturalness. This work represents a significant advancement in Hakka language technology and provides a valuable resource for language preservation and revitalization efforts.

Bjoern List, Li-Wei Chen, Kartik Bali et al.

Unrolling training trajectories over time strongly influences the inference accuracy of neural network-augmented physics simulators. We analyze this in three variants of training neural time-steppers. In addition to one-step setups and fully differentiable unrolling, we include a third, less widely used variant: unrolling without temporal gradients. Comparing networks trained with these three modalities disentangles the two dominant effects of unrolling, training distribution shift and long-term gradients. We present detailed study across physical systems, network sizes and architectures, training setups, and test scenarios. It also encompasses two simulation modes: In prediction setups, we rely solely on neural networks to compute a trajectory. In contrast, correction setups include a numerical solver that is supported by a neural network. Spanning these variations, our study provides the empirical basis for our main findings: Non-differentiable but unrolled training with a numerical solver in a correction setup can yield substantial improvements over a fully differentiable prediction setup not utilizing this solver. The accuracy of models trained in a fully differentiable setup differs compared to their non-differentiable counterparts. Differentiable ones perform best in a comparison among correction networks as well as among prediction setups. For both, the accuracy of non-differentiable unrolling comes close. Furthermore, we show that these behaviors are invariant to the physical system, the network architecture and size, and the numerical scheme. These results motivate integrating non-differentiable numerical simulators into training setups even if full differentiability is unavailable. We show the convergence rate of common architectures to be low compared to numerical algorithms. This motivates correction setups combining neural and numerical parts which utilize benefits of both.

Kuan-Tang Huang, Li-Wei Chen, Hung-Shin Lee et al.

Pre-trained automatic speech recognition (ASR) models have demonstrated strong performance on a variety of tasks. However, their performance can degrade substantially when the input audio comes from different recording channels. While previous studies have demonstrated this phenomenon, it is often attributed to the mismatch between training and testing corpora. This study argues that variations in speech characteristics caused by different recording channels can fundamentally harm ASR performance. To address this limitation, we propose a normalization technique designed to mitigate the impact of channel variation by aligning internal feature representations in the ASR model with those derived from a clean reference channel. This approach significantly improves ASR performance on previously unseen channels and languages, highlighting its ability to generalize across channel and language differences.

Hyung Gun Chi, Zakaria Aldeneh, Tatiana Likhomanenko et al. · apple-ml

We introduce DiceHuBERT, a knowledge distillation framework for compressing HuBERT, a widely used self-supervised learning (SSL)-based speech foundation model. Unlike existing distillation methods that rely on layer-wise and feature-wise mapping between teacher and student models, DiceHuBERT leverages HuBERT's iterative self-distillation mechanism by directly replacing the original model with a student model. This replacement allows the student to be trained using the same SSL objective used when pre-training HuBERT, eliminating the need for additional modules or architectural constraints. Experimental results on SUPERB show that DiceHuBERT consistently outperforms existing distillation methods, improving phoneme recognition performance by over 21% and ASR performance by more than 14%. Furthermore, DiceHuBERT demonstrates competitive performance across multiple tasks, highlighting its clear advantage.

Ta-Chung Chi, Ting-Han Fan, Li-Wei Chen et al.

The use of positional embeddings in transformer language models is widely accepted. However, recent research has called into question the necessity of such embeddings. We further extend this inquiry by demonstrating that a randomly initialized and frozen transformer language model, devoid of positional embeddings, inherently encodes strong positional information through the shrinkage of self-attention variance. To quantify this variance, we derive the underlying distribution of each step within a transformer layer. Through empirical validation using a fully pretrained model, we show that the variance shrinkage effect still persists after extensive gradient updates. Our findings serve to justify the decision to discard positional embeddings and thus facilitate more efficient pretraining of transformer language models.

Björn List, Li-Wei Chen, Nils Thuerey

In this paper, we train turbulence models based on convolutional neural networks. These learned turbulence models improve under-resolved low resolution solutions to the incompressible Navier-Stokes equations at simulation time. Our study involves the development of a differentiable numerical solver that supports the propagation of optimisation gradients through multiple solver steps. The significance of this property is demonstrated by the superior stability and accuracy of those models that unroll more solver steps during training. Furthermore, we introduce loss terms based on turbulence physics that further improve the model accuracy. This approach is applied to three two-dimensional turbulence flow scenarios, a homogeneous decaying turbulence case, a temporally evolving mixing layer, and a spatially evolving mixing layer. Our models achieve significant improvements of long-term a-posteriori statistics when compared to no-model simulations, without requiring these statistics to be directly included in the learning targets. At inference time, our proposed method also gains substantial performance improvements over similarly accurate, purely numerical methods.

Georg Kohl, Li-Wei Chen, Nils Thuerey

Simulations that produce three-dimensional data are ubiquitous in science, ranging from fluid flows to plasma physics. We propose a similarity model based on entropy, which allows for the creation of physically meaningful ground truth distances for the similarity assessment of scalar and vectorial data, produced from transport and motion-based simulations. Utilizing two data acquisition methods derived from this model, we create collections of fields from numerical PDE solvers and existing simulation data repositories. Furthermore, a multiscale CNN architecture that computes a volumetric similarity metric (VolSiM) is proposed. To the best of our knowledge this is the first learning method inherently designed to address the challenges arising for the similarity assessment of high-dimensional simulation data. Additionally, the tradeoff between a large batch size and an accurate correlation computation for correlation-based loss functions is investigated, and the metric's invariance with respect to rotation and scale operations is analyzed. Finally, the robustness and generalization of VolSiM is evaluated on a large range of test data, as well as a particularly challenging turbulence case study, that is close to potential real-world applications.

Li-Wei Chen, Alexander Rudnicky

While Wav2Vec 2.0 has been proposed for speech recognition (ASR), it can also be used for speech emotion recognition (SER); its performance can be significantly improved using different fine-tuning strategies. Two baseline methods, vanilla fine-tuning (V-FT) and task adaptive pretraining (TAPT) are first presented. We show that V-FT is able to outperform state-of-the-art models on the IEMOCAP dataset. TAPT, an existing NLP fine-tuning strategy, further improves the performance on SER. We also introduce a novel fine-tuning method termed P-TAPT, which modifies the TAPT objective to learn contextualized emotion representations. Experiments show that P-TAPT performs better than TAPT, especially under low-resource settings. Compared to prior works in this literature, our top-line system achieved a 7.4\% absolute improvement in unweighted accuracy (UA) over the state-of-the-art performance on IEMOCAP. Our code is publicly available.

Li-Wei Chen, Alexander Rudnicky

In this paper, we present a novel architecture to realize fine-grained style control on the transformer-based text-to-speech synthesis (TransformerTTS). Specifically, we model the speaking style by extracting a time sequence of local style tokens (LST) from the reference speech. The existing content encoder in TransformerTTS is then replaced by our designed cross-attention blocks for fusion and alignment between content and style. As the fusion is performed along with the skip connection, our cross-attention block provides a good inductive bias to gradually infuse the phoneme representation with a given style. Additionally, we prevent the style embedding from encoding linguistic content by randomly truncating LST during training and using wav2vec 2.0 features. Experiments show that with fine-grained style control, our system performs better in terms of naturalness, intelligibility, and style transferability. Our code and samples are publicly available.

Li-Wei Chen, Nils Thuerey

The present study investigates the accurate inference of Reynolds-averaged Navier-Stokes solutions for the compressible flow over aerofoils in two dimensions with a deep neural network. Our approach yields networks that learn to generate precise flow fields for varying body-fitted, structured grids by providing them with an encoding of the corresponding mapping to a canonical space for the solutions. We apply the deep neural network model to a benchmark case of incompressible flow at randomly given angles of attack and Reynolds numbers and achieve an improvement of more than an order of magnitude compared to previous work. Further, for transonic flow cases, the deep neural network model accurately predicts complex flow behaviour at high Reynolds numbers, such as shock wave/boundary layer interaction, and quantitative distributions like pressure coefficient, skin friction coefficient as well as wake total pressure profiles downstream of aerofoils. The proposed deep learning method significantly speeds up the predictions of flow fields and shows promise for enabling fast aerodynamic designs.

Takashi Maekaku, Xuankai Chang, Yuya Fujita et al.

We present a system for the Zero Resource Speech Challenge 2021, which combines a Contrastive Predictive Coding (CPC) with deep cluster. In deep cluster, we first prepare pseudo-labels obtained by clustering the outputs of a CPC network with k-means. Then, we train an additional autoregressive model to classify the previously obtained pseudo-labels in a supervised manner. Phoneme discriminative representation is achieved by executing the second-round clustering with the outputs of the final layer of the autoregressive model. We show that replacing a Transformer layer with a Conformer layer leads to a further gain in a lexical metric. Experimental results show that a relative improvement of 35% in a phonetic metric, 1.5% in the lexical metric, and 2.3% in a syntactic metric are achieved compared to a baseline method of CPC-small which is trained on LibriSpeech 460h data. We achieve top results in this challenge with the syntactic metric.

Li-Wei Chen, Wei-Chen Chiu, Chin-Tien Wu

It is well known that semantic segmentation neural networks (SSNNs) produce dense segmentation maps to resolve the objects' boundaries while restrict the prediction on down-sampled grids to alleviate the computational cost. A striking balance between the accuracy and the training cost of the SSNNs such as U-Net exists. We propose a spectral analysis to investigate the correlations among the resolution of the down sampled grid, the loss function and the accuracy of the SSNNs. By analyzing the network back-propagation process in frequency domain, we discover that the traditional loss function, cross-entropy, and the key features of CNN are mainly affected by the low-frequency components of segmentation labels. Our discoveries can be applied to SSNNs in several ways including (i) determining an efficient low resolution grid for resolving the segmentation maps (ii) pruning the networks by truncating the high frequency decoder features for saving computation costs, and (iii) using block-wise weak annotation for saving the labeling time. Experimental results shown in this paper agree with our spectral analysis for the networks such as DeepLab V3+ and Deep Aggregation Net (DAN).

Li-Wei Chen, Hung-Yi Lee, Yu Tsao

This paper focuses on using voice conversion (VC) to improve the speech intelligibility of surgical patients who have had parts of their articulators removed. Due to the difficulty of data collection, VC without parallel data is highly desired. Although techniques for unparallel VC, for example, CycleGAN, have been developed, they usually focus on transforming the speaker identity, and directly transforming the speech of one speaker to that of another speaker and as such do not address the task here. In this paper, we propose a new approach for unparallel VC. The proposed approach transforms impaired speech to normal speech while preserving the linguistic content and speaker characteristics. To our knowledge, this is the first end-to-end GAN-based unsupervised VC model applied to impaired speech. The experimental results show that the proposed approach outperforms CycleGAN.