Lin-shan Lee

Guan-Ting Lin, Yung-Sung Chuang, Ho-Lam Chung et al. · meta-ai, mit

Spoken Question Answering (SQA) is to find the answer from a spoken document given a question, which is crucial for personal assistants when replying to the queries from the users. Existing SQA methods all rely on Automatic Speech Recognition (ASR) transcripts. Not only does ASR need to be trained with massive annotated data that are time and cost-prohibitive to collect for low-resourced languages, but more importantly, very often the answers to the questions include name entities or out-of-vocabulary words that cannot be recognized correctly. Also, ASR aims to minimize recognition errors equally over all words, including many function words irrelevant to the SQA task. Therefore, SQA without ASR transcripts (textless) is always highly desired, although known to be very difficult. This work proposes Discrete Spoken Unit Adaptive Learning (DUAL), leveraging unlabeled data for pre-training and fine-tuned by the SQA downstream task. The time intervals of spoken answers can be directly predicted from spoken documents. We also release a new SQA benchmark corpus, NMSQA, for data with more realistic scenarios. We empirically showed that DUAL yields results comparable to those obtained by cascading ASR and text QA model and robust to real-world data. Our code and model will be open-sourced.

Liang-Hsuan Tseng, En-Pei Hu, Cheng-Han Chiang et al.

Unsupervised automatic speech recognition (ASR) aims to learn the mapping between the speech signal and its corresponding textual transcription without the supervision of paired speech-text data. A word/phoneme in the speech signal is represented by a segment of speech signal with variable length and unknown boundary, and this segmental structure makes learning the mapping between speech and text challenging, especially without paired data. In this paper, we propose REBORN,Reinforcement-Learned Boundary Segmentation with Iterative Training for Unsupervised ASR. REBORN alternates between (1) training a segmentation model that predicts the boundaries of the segmental structures in speech signals and (2) training the phoneme prediction model, whose input is the speech feature segmented by the segmentation model, to predict a phoneme transcription. Since supervised data for training the segmentation model is not available, we use reinforcement learning to train the segmentation model to favor segmentations that yield phoneme sequence predictions with a lower perplexity. We conduct extensive experiments and find that under the same setting, REBORN outperforms all prior unsupervised ASR models on LibriSpeech, TIMIT, and five non-English languages in Multilingual LibriSpeech. We comprehensively analyze why the boundaries learned by REBORN improve the unsupervised ASR performance.

Chyi-Jiunn Lin, Guan-Ting Lin, Yung-Sung Chuang et al.

Spoken Question Answering (SQA) is essential for machines to reply to user's question by finding the answer span within a given spoken passage. SQA has been previously achieved without ASR to avoid recognition errors and Out-of-Vocabulary (OOV) problems. However, the real-world problem of Open-domain SQA (openSQA), in which the machine needs to first retrieve passages that possibly contain the answer from a spoken archive in addition, was never considered. This paper proposes the first known end-to-end framework, Speech Dense Passage Retriever (SpeechDPR), for the retrieval component of the openSQA problem. SpeechDPR learns a sentence-level semantic representation by distilling knowledge from the cascading model of unsupervised ASR (UASR) and text dense retriever (TDR). No manually transcribed speech data is needed. Initial experiments showed performance comparable to the cascading model of UASR and TDR, and significantly better when UASR was poor, verifying this approach is more robust to speech recognition errors.

Heng-Jui Chang, Hung-yi Lee, Lin-shan Lee

Automatic speech recognition (ASR) technologies today are primarily optimized for given datasets; thus, any changes in the application environment (e.g., acoustic conditions or topic domains) may inevitably degrade the performance. We can collect new data describing the new environment and fine-tune the system, but this naturally leads to higher error rates for the earlier datasets, referred to as catastrophic forgetting. The concept of lifelong learning (LLL) aiming to enable a machine to sequentially learn new tasks from new datasets describing the changing real world without forgetting the previously learned knowledge is thus brought to attention. This paper reports, to our knowledge, the first effort to extensively consider and analyze the use of various approaches of LLL in end-to-end (E2E) ASR, including proposing novel methods in saving data for past domains to mitigate the catastrophic forgetting problem. An overall relative reduction of 28.7% in WER was achieved compared to the fine-tuning baseline when sequentially learning on three very different benchmark corpora. This can be the first step toward the highly desired ASR technologies capable of synchronizing with the continuously changing real world.

Yist Y. Lin, Chung-Ming Chien, Jheng-Hao Lin et al.

Any-to-any voice conversion aims to convert the voice from and to any speakers even unseen during training, which is much more challenging compared to one-to-one or many-to-many tasks, but much more attractive in real-world scenarios. In this paper we proposed FragmentVC, in which the latent phonetic structure of the utterance from the source speaker is obtained from Wav2Vec 2.0, while the spectral features of the utterance(s) from the target speaker are obtained from log mel-spectrograms. By aligning the hidden structures of the two different feature spaces with a two-stage training process, FragmentVC is able to extract fine-grained voice fragments from the target speaker utterance(s) and fuse them into the desired utterance, all based on the attention mechanism of Transformer as verified with analysis on attention maps, and is accomplished end-to-end. This approach is trained with reconstruction loss only without any disentanglement considerations between content and speaker information and doesn't require parallel data. Objective evaluation based on speaker verification and subjective evaluation with MOS both showed that this approach outperformed SOTA approaches, such as AdaIN-VC and AutoVC.

Chien-yu Huang, Yist Y. Lin, Hung-yi Lee et al.

Substantial improvements have been achieved in recent years in voice conversion, which converts the speaker characteristics of an utterance into those of another speaker without changing the linguistic content of the utterance. Nonetheless, the improved conversion technologies also led to concerns about privacy and authentication. It thus becomes highly desired to be able to prevent one's voice from being improperly utilized with such voice conversion technologies. This is why we report in this paper the first known attempt to perform adversarial attack on voice conversion. We introduce human imperceptible noise into the utterances of a speaker whose voice is to be defended. Given these adversarial examples, voice conversion models cannot convert other utterances so as to sound like being produced by the defended speaker. Preliminary experiments were conducted on two currently state-of-the-art zero-shot voice conversion models. Objective and subjective evaluation results in both white-box and black-box scenarios are reported. It was shown that the speaker characteristics of the converted utterances were made obviously different from those of the defended speaker, while the adversarial examples of the defended speaker are not distinguishable from the authentic utterances.

Heng-Jui Chang, Alexander H. Liu, Hung-yi Lee et al.

Whispering is an important mode of human speech, but no end-to-end recognition results for it were reported yet, probably due to the scarcity of available whispered speech data. In this paper, we present several approaches for end-to-end (E2E) recognition of whispered speech considering the special characteristics of whispered speech and the scarcity of data. This includes a frequency-weighted SpecAugment policy and a frequency-divided CNN feature extractor for better capturing the high-frequency structures of whispered speech, and a layer-wise transfer learning approach to pre-train a model with normal or normal-to-whispered converted speech then fine-tune it with whispered speech to bridge the gap between whispered and normal speech. We achieve an overall relative reduction of 19.8% in PER and 44.4% in CER on a relatively small whispered TIMIT corpus. The results indicate as long as we have a good E2E model pre-trained on normal or pseudo-whispered speech, a relatively small set of whispered speech may suffice to obtain a reasonably good E2E whispered speech recognizer.

Alexander H. Liu, Tzu-Wei Sung, Shun-Po Chuang et al.

In this paper, we investigate the benefit that off-the-shelf word embedding can bring to the sequence-to-sequence (seq-to-seq) automatic speech recognition (ASR). We first introduced the word embedding regularization by maximizing the cosine similarity between a transformed decoder feature and the target word embedding. Based on the regularized decoder, we further proposed the fused decoding mechanism. This allows the decoder to consider the semantic consistency during decoding by absorbing the information carried by the transformed decoder feature, which is learned to be close to the target word embedding. Initial results on LibriSpeech demonstrated that pre-trained word embedding can significantly lower ASR recognition error with a negligible cost, and the choice of word embedding algorithms among Skip-gram, CBOW and BERT is important.

Alexander H. Liu, Tao Tu, Hung-yi Lee et al.

In this paper we propose a Sequential Representation Quantization AutoEncoder (SeqRQ-AE) to learn from primarily unpaired audio data and produce sequences of representations very close to phoneme sequences of speech utterances. This is achieved by proper temporal segmentation to make the representations phoneme-synchronized, and proper phonetic clustering to have total number of distinct representations close to the number of phonemes. Mapping between the distinct representations and phonemes is learned from a small amount of annotated paired data. Preliminary experiments on LJSpeech demonstrated the learned representations for vowels have relative locations in latent space in good parallel to that shown in the IPA vowel chart defined by linguistics experts. With less than 20 minutes of annotated speech, our method outperformed existing methods on phoneme recognition and is able to synthesize intelligible speech that beats our baseline model.

Gene-Ping Yang, Szu-Lin Wu, Yao-Wen Mao et al.

Permutation Invariant Training (PIT) has long been a stepping stone method for training speech separation model in handling the label ambiguity problem. With PIT selecting the minimum cost label assignments dynamically, very few studies considered the separation problem to be optimizing both the model parameters and the label assignments, but focused on searching for good model architecture and parameters. In this paper, we investigate instead for a given model architecture the various flexible label assignment strategies for training the model, rather than directly using PIT. Surprisingly, we discover a significant performance boost compared to PIT is possible if the model is trained with fixed label assignments and a good set of labels is chosen. With fixed label training cascaded between two sections of PIT, we achieved the state-of-the-art performance on WSJ0-2mix without changing the model architecture at all.

Yung-Sung Chuang, Chi-Liang Liu, Hung-Yi Lee et al.

While various end-to-end models for spoken language understanding tasks have been explored recently, this paper is probably the first known attempt to challenge the very difficult task of end-to-end spoken question answering (SQA). Learning from the very successful BERT model for various text processing tasks, here we proposed an audio-and-text jointly learned SpeechBERT model. This model outperformed the conventional approach of cascading ASR with the following text question answering (TQA) model on datasets including ASR errors in answer spans, because the end-to-end model was shown to be able to extract information out of audio data before ASR produced errors. When ensembling the proposed end-to-end model with the cascade architecture, even better performance was achieved. In addition to the potential of end-to-end SQA, the SpeechBERT can also be considered for many other spoken language understanding tasks just as BERT for many text processing tasks.

Gene-Ping Yang, Chao-I Tuan, Hung-Yi Lee et al.

Speech separation has been very successful with deep learning techniques. Substantial effort has been reported based on approaches over spectrogram, which is well known as the standard time-and-frequency cross-domain representation for speech signals. It is highly correlated to the phonetic structure of speech, or "how the speech sounds" when perceived by human, but primarily frequency domain features carrying temporal behaviour. Very impressive work achieving speech separation over time domain was reported recently, probably because waveforms in time domain may describe the different realizations of speech in a more precise way than spectrogram. In this paper, we propose a framework properly integrating the above two directions, hoping to achieve both purposes. We construct a time-and-frequency feature map by concatenating the 1-dim convolution encoded feature map (for time domain) and the spectrogram (for frequency domain), which was then processed by an embedding network and clustering approaches very similar to those used in time and frequency domain prior works. In this way, the information in the time and frequency domains, as well as the interactions between them, can be jointly considered during embedding and clustering. Very encouraging results (state-of-the-art to our knowledge) were obtained with WSJ0-2mix dataset in preliminary experiments.

Yi-Chen Chen, Sung-Feng Huang, Hung-yi Lee et al.

Producing a large amount of annotated speech data for training ASR systems remains difficult for more than 95% of languages all over the world which are low-resourced. However, we note human babies start to learn the language by the sounds (or phonetic structures) of a small number of exemplar words, and "generalize" such knowledge to other words without hearing a large amount of data. We initiate some preliminary work in this direction. Audio Word2Vec is used to learn the phonetic structures from spoken words (signal segments), while another autoencoder is used to learn the phonetic structures from text words. The relationships among the above two can be learned jointly, or separately after the above two are well trained. This relationship can be used in speech recognition with very low resource. In the initial experiments on the TIMIT dataset, only 2.1 hours of speech data (in which 2500 spoken words were annotated and the rest unlabeled) gave a word error rate of 44.6%, and this number can be reduced to 34.2% if 4.1 hr of speech data (in which 20000 spoken words were annotated) were given. These results are not satisfactory, but a good starting point.

Kuan-Yu Chen, Che-Ping Tsai, Da-Rong Liu et al.

Producing a large annotated speech corpus for training ASR systems remains difficult for more than 95% of languages all over the world which are low-resourced, but collecting a relatively big unlabeled data set for such languages is more achievable. This is why some initial effort have been reported on completely unsupervised speech recognition learned from unlabeled data only, although with relatively high error rates. In this paper, we develop a Generative Adversarial Network (GAN) to achieve this purpose, in which a Generator and a Discriminator learn from each other iteratively to improve the performance. We further use a set of Hidden Markov Models (HMMs) iteratively refined from the machine generated labels to work in harmony with the GAN. The initial experiments on TIMIT data set achieve an phone error rate of 33.1%, which is 8.5% lower than the previous state-of-the-art.

Sung-Feng Huang, Yi-Chen Chen, Hung-yi Lee et al.

Embedding audio signal segments into vectors with fixed dimensionality is attractive because all following processing will be easier and more efficient, for example modeling, classifying or indexing. Audio Word2Vec previously proposed was shown to be able to represent audio segments for spoken words as such vectors carrying information about the phonetic structures of the signal segments. However, each linguistic unit (word, syllable, phoneme in text form) corresponds to unlimited number of audio segments with vector representations inevitably spread over the embedding space, which causes some confusion. It is therefore desired to better cluster the audio embeddings such that those corresponding to the same linguistic unit can be more compactly distributed. In this paper, inspired by Siamese networks, we propose some approaches to achieve the above goal. This includes identifying positive and negative pairs from unlabeled data for Siamese style training, disentangling acoustic factors such as speaker characteristics from the audio embedding, handling unbalanced data distribution, and having the embedding processes learn from the adjacency relationships among data points. All these can be done in an unsupervised way. Improved performance was obtained in preliminary experiments on the LibriSpeech data set, including clustering characteristics analysis and applications of spoken term detection.

Alexander H. Liu, Hung-yi Lee, Lin-shan Lee

In this paper we proposed a novel Adversarial Training (AT) approach for end-to-end speech recognition using a Criticizing Language Model (CLM). In this way the CLM and the automatic speech recognition (ASR) model can challenge and learn from each other iteratively to improve the performance. Since the CLM only takes the text as input, huge quantities of unpaired text data can be utilized in this approach within end-to-end training. Moreover, AT can be applied to any end-to-end ASR model using any deep-learning-based language modeling frameworks, and compatible with any existing end-to-end decoding method. Initial results with an example experimental setup demonstrated the proposed approach is able to gain consistent improvements efficiently from auxiliary text data under different scenarios.

Yi-Chen Chen, Chia-Hao Shen, Sung-Feng Huang et al.

Producing a large amount of annotated speech data for training ASR systems remains difficult for more than 95% of languages all over the world which are low-resourced. However, we note human babies start to learn the language by the sounds of a small number of exemplar words without hearing a large amount of data. We initiate some preliminary work in this direction in this paper. Audio Word2Vec is used to obtain embeddings of spoken words which carry phonetic information extracted from the signals. An autoencoder is used to generate embeddings of text words based on the articulatory features for the phoneme sequences. Both sets of embeddings for spoken and text words describe similar phonetic structures among words in their respective latent spaces. A mapping relation from the audio embeddings to text embeddings actually gives the word-level ASR. This can be learned by aligning a small number of spoken words and the corresponding text words in the embedding spaces. In the initial experiments only 200 annotated spoken words and one hour of speech data without annotation gave a word accuracy of 27.5%, which is low but a good starting point.

Cheng-chieh Yeh, Po-chun Hsu, Ju-chieh Chou et al.

Speaking rate refers to the average number of phonemes within some unit time, while the rhythmic patterns refer to duration distributions for realizations of different phonemes within different phonetic structures. Both are key components of prosody in speech, which is different for different speakers. Models like cycle-consistent adversarial network (Cycle-GAN) and variational auto-encoder (VAE) have been successfully applied to voice conversion tasks without parallel data. However, due to the neural network architectures and feature vectors chosen for these approaches, the length of the predicted utterance has to be fixed to that of the input utterance, which limits the flexibility in mimicking the speaking rates and rhythmic patterns for the target speaker. On the other hand, sequence-to-sequence learning model was used to remove the above length constraint, but parallel training data are needed. In this paper, we propose an approach utilizing sequence-to-sequence model trained with unsupervised Cycle-GAN to perform the transformation between the phoneme posteriorgram sequences for different speakers. In this way, the length constraint mentioned above is removed to offer rhythm-flexible voice conversion without requiring parallel data. Preliminary evaluation on two datasets showed very encouraging results.

Yu-Hsuan Wang, Hung-yi Lee, Lin-shan Lee

While Word2Vec represents words (in text) as vectors carrying semantic information, audio Word2Vec was shown to be able to represent signal segments of spoken words as vectors carrying phonetic structure information. Audio Word2Vec can be trained in an unsupervised way from an unlabeled corpus, except the word boundaries are needed. In this paper, we extend audio Word2Vec from word-level to utterance-level by proposing a new segmental audio Word2Vec, in which unsupervised spoken word boundary segmentation and audio Word2Vec are jointly learned and mutually enhanced, so an utterance can be directly represented as a sequence of vectors carrying phonetic structure information. This is achieved by a segmental sequence-to-sequence autoencoder (SSAE), in which a segmentation gate trained with reinforcement learning is inserted in the encoder. Experiments on English, Czech, French and German show very good performance in both unsupervised spoken word segmentation and spoken term detection applications (significantly better than frame-based DTW).

Yi-Chen Chen, Sung-Feng Huang, Chia-Hao Shen et al.

Word embedding or Word2Vec has been successful in offering semantics for text words learned from the context of words. Audio Word2Vec was shown to offer phonetic structures for spoken words (signal segments for words) learned from signals within spoken words. This paper proposes a two-stage framework to perform phonetic-and-semantic embedding on spoken words considering the context of the spoken words. Stage 1 performs phonetic embedding with speaker characteristics disentangled. Stage 2 then performs semantic embedding in addition. We further propose to evaluate the phonetic-and-semantic nature of the audio embeddings obtained in Stage 2 by parallelizing with text embeddings. In general, phonetic structure and semantics inevitably disturb each other. For example the words "brother" and "sister" are close in semantics but very different in phonetic structure, while the words "brother" and "bother" are in the other way around. But phonetic-and-semantic embedding is attractive, as shown in the initial experiments on spoken document retrieval. Not only spoken documents including the spoken query can be retrieved based on the phonetic structures, but spoken documents semantically related to the query but not including the query can also be retrieved based on the semantics.

Che-Ping Tsai, Yi-Lin Tuan, Lin-shan Lee

Spoken content processing (such as retrieval and browsing) is maturing, but the singing content is still almost completely left out. Songs are human voice carrying plenty of semantic information just as speech, and may be considered as a special type of speech with highly flexible prosody. The various problems in song audio, for example the significantly changing phone duration over highly flexible pitch contours, make the recognition of lyrics from song audio much more difficult. This paper reports an initial attempt towards this goal. We collected music-removed version of English songs directly from commercial singing content. The best results were obtained by TDNN-LSTM with data augmentation with 3-fold speed perturbation plus some special approaches. The WER achieved (73.90%) was significantly lower than the baseline (96.21%), but still relatively high.

Ju-chieh Chou, Cheng-chieh Yeh, Hung-yi Lee et al.

Recently, cycle-consistent adversarial network (Cycle-GAN) has been successfully applied to voice conversion to a different speaker without parallel data, although in those approaches an individual model is needed for each target speaker. In this paper, we propose an adversarial learning framework for voice conversion, with which a single model can be trained to convert the voice to many different speakers, all without parallel data, by separating the speaker characteristics from the linguistic content in speech signals. An autoencoder is first trained to extract speaker-independent latent representations and speaker embedding separately using another auxiliary speaker classifier to regularize the latent representation. The decoder then takes the speaker-independent latent representation and the target speaker embedding as the input to generate the voice of the target speaker with the linguistic content of the source utterance. The quality of decoder output is further improved by patching with the residual signal produced by another pair of generator and discriminator. A target speaker set size of 20 was tested in the preliminary experiments, and very good voice quality was obtained. Conventional voice conversion metrics are reported. We also show that the speaker information has been properly reduced from the latent representations.

Chih-Wei Lee, Yau-Shian Wang, Tsung-Yuan Hsu et al.

Conventional seq2seq chatbot models only try to find the sentences with the highest probabilities conditioned on the input sequences, without considering the sentiment of the output sentences. Some research works trying to modify the sentiment of the output sequences were reported. In this paper, we propose five models to scale or adjust the sentiment of the chatbot response: persona-based model, reinforcement learning, plug and play model, sentiment transformation network and cycleGAN, all based on the conventional seq2seq model. We also develop two evaluation metrics to estimate if the responses are reasonable given the input. These metrics together with other two popularly used metrics were used to analyze the performance of the five proposed models on different aspects, and reinforcement learning and cycleGAN were shown to be very attractive. The evaluation metrics were also found to be well correlated with human evaluation.

Da-Rong Liu, Kuan-Yu Chen, Hung-Yi Lee et al.

Unsupervised discovery of acoustic tokens from audio corpora without annotation and learning vector representations for these tokens have been widely studied. Although these techniques have been shown successful in some applications such as query-by-example Spoken Term Detection (STD), the lack of mapping relationships between these discovered tokens and real phonemes have limited the down-stream applications. This paper represents probably the first attempt towards the goal of completely unsupervised phoneme recognition, or mapping audio signals to phoneme sequences without phoneme-labeled audio data. The basic idea is to cluster the embedded acoustic tokens and learn the mapping between the cluster sequences and the unknown phoneme sequences with a Generative Adversarial Network (GAN). An unsupervised phoneme recognition accuracy of 36% was achieved in the preliminary experiments.

Cheng-Tao Chung, Lin-Shan Lee

In this paper, we compare two paradigms for unsupervised discovery of structured acoustic tokens directly from speech corpora without any human annotation. The Multigranular Paradigm seeks to capture all available information in the corpora with multiple sets of tokens for different model granularities. The Hierarchical Paradigm attempts to jointly learn several levels of signal representations in a hierarchical structure. The two paradigms are unified within a theoretical framework in this paper. Query-by-Example Spoken Term Detection (QbE-STD) experiments on the QUESST dataset of MediaEval 2015 verifies the competitiveness of the acoustic tokens. The Enhanced Relevance Score (ERS) proposed in this work improves both paradigms for the task of QbE-STD. We also list results on the ABX evaluation task of the Zero Resource Challenge 2015 for comparison of the Paradigms.

Zih-Wei Lin, Tzu-Wei Sung, Hung-Yi Lee et al.

Distributed word representations have been shown to be very useful in various natural language processing (NLP) application tasks. These word vectors learned from huge corpora very often carry both semantic and syntactic information of words. However, it is well known that each individual user has his own language patterns because of different factors such as interested topics, friend groups, social activities, wording habits, etc., which may imply some kind of personalized semantics. With such personalized semantics, the same word may imply slightly differently for different users. For example, the word "Cappuccino" may imply "Leisure", "Joy", "Excellent" for a user enjoying coffee, by only a kind of drink for someone else. Such personalized semantics of course cannot be carried by the standard universal word vectors trained with huge corpora produced by many people. In this paper, we propose a framework to train different personalized word vectors for different users based on the very successful continuous skip-gram model using the social network data posted by many individual users. In this framework, universal background word vectors are first learned from the background corpora, and then adapted by the personalized corpus for each individual user to learn the personalized word vectors. We use two application tasks to evaluate the quality of the personalized word vectors obtained in this way, the user prediction task and the sentence completion task. These personalized word vectors were shown to carry some personalized semantics and offer improved performance on these two evaluation tasks.

Bo-Ru Lu, Frank Shyu, Yun-Nung Chen et al.

Connectionist temporal classification (CTC) is a powerful approach for sequence-to-sequence learning, and has been popularly used in speech recognition. The central ideas of CTC include adding a label "blank" during training. With this mechanism, CTC eliminates the need of segment alignment, and hence has been applied to various sequence-to-sequence learning problems. In this work, we applied CTC to abstractive summarization for spoken content. The "blank" in this case implies the corresponding input data are less important or noisy; thus it can be ignored. This approach was shown to outperform the existing methods in term of ROUGE scores over Chinese Gigaword and MATBN corpora. This approach also has the nice property that the ordering of words or characters in the input documents can be better preserved in the generated summaries.

Cheng-Tao Chung, Cheng-Yu Tsai, Chia-Hsiang Liu et al.

In this paper we aim to automatically discover high quality frame-level speech features and acoustic tokens directly from unlabeled speech data. A Multi-granular Acoustic Tokenizer (MAT) was proposed for automatic discovery of multiple sets of acoustic tokens from the given corpus. Each acoustic token set is specified by a set of hyperparameters describing the model configuration. These different sets of acoustic tokens carry different characteristics for the given corpus and the language behind, thus can be mutually reinforced. The multiple sets of token labels are then used as the targets of a Multi-target Deep Neural Network (MDNN) trained on frame-level acoustic features. Bottleneck features extracted from the MDNN are then used as the feedback input to the MAT and the MDNN itself in the next iteration. The multi-granular acoustic token sets and the frame-level speech features can be iteratively optimized in the iterative deep learning framework. We call this framework the Multi-granular Acoustic Tokenizing Deep Neural Network (MATDNN). The results were evaluated using the metrics and corpora defined in the Zero Resource Speech Challenge organized at Interspeech 2015, and improved performance was obtained with a set of experiments of query-by-example spoken term detection on the same corpora. Visualization for the discovered tokens against the English phonemes was also shown.

Cheng-Kuan Wei, Cheng-Tao Chung, Hung-Yi Lee et al.

It is well known that recognizers personalized to each user are much more effective than user-independent recognizers. With the popularity of smartphones today, although it is not difficult to collect a large set of audio data for each user, it is difficult to transcribe it. However, it is now possible to automatically discover acoustic tokens from unlabeled personal data in an unsupervised way. We therefore propose a multi-task deep learning framework called a phoneme-token deep neural network (PTDNN), jointly trained from unsupervised acoustic tokens discovered from unlabeled data and very limited transcribed data for personalized acoustic modeling. We term this scenario "weakly supervised". The underlying intuition is that the high degree of similarity between the HMM states of acoustic token models and phoneme models may help them learn from each other in this multi-task learning framework. Initial experiments performed over a personalized audio data set recorded from Facebook posts demonstrated that very good improvements can be achieved in both frame accuracy and word accuracy over popularly-considered baselines such as fDLR, speaker code and lightly supervised adaptation. This approach complements existing speaker adaptation approaches and can be used jointly with such techniques to yield improved results.

Lang-Chi Yu, Hung-yi Lee, Lin-shan Lee

Headline generation for spoken content is important since spoken content is difficult to be shown on the screen and browsed by the user. It is a special type of abstractive summarization, for which the summaries are generated word by word from scratch without using any part of the original content. Many deep learning approaches for headline generation from text document have been proposed recently, all requiring huge quantities of training data, which is difficult for spoken document summarization. In this paper, we propose an ASR error modeling approach to learn the underlying structure of ASR error patterns and incorporate this model in an Attentive Recurrent Neural Network (ARNN) architecture. In this way, the model for abstractive headline generation for spoken content can be learned from abundant text data and the ASR data for some recognizers. Experiments showed very encouraging results and verified that the proposed ASR error model works well even when the input spoken content is recognized by a recognizer very different from the one the model learned from.

Yen-Chen Wu, Tzu-Hsiang Lin, Yang-De Chen et al.

User-machine interaction is important for spoken content retrieval. For text content retrieval, the user can easily scan through and select on a list of retrieved item. This is impossible for spoken content retrieval, because the retrieved items are difficult to show on screen. Besides, due to the high degree of uncertainty for speech recognition, the retrieval results can be very noisy. One way to counter such difficulties is through user-machine interaction. The machine can take different actions to interact with the user to obtain better retrieval results before showing to the user. The suitable actions depend on the retrieval status, for example requesting for extra information from the user, returning a list of topics for user to select, etc. In our previous work, some hand-crafted states estimated from the present retrieval results are used to determine the proper actions. In this paper, we propose to use Deep-Q-Learning techniques instead to determine the machine actions for interactive spoken content retrieval. Deep-Q-Learning bypasses the need for estimation of the hand-crafted states, and directly determine the best action base on the present retrieval status even without any human knowledge. It is shown to achieve significantly better performance compared with the previous hand-crafted states.

Wei Fang, Jui-Yang Hsu, Hung-yi Lee et al.

Multimedia or spoken content presents more attractive information than plain text content, but the former is more difficult to display on a screen and be selected by a user. As a result, accessing large collections of the former is much more difficult and time-consuming than the latter for humans. It's therefore highly attractive to develop machines which can automatically understand spoken content and summarize the key information for humans to browse over. In this endeavor, a new task of machine comprehension of spoken content was proposed recently. The initial goal was defined as the listening comprehension test of TOEFL, a challenging academic English examination for English learners whose native languages are not English. An Attention-based Multi-hop Recurrent Neural Network (AMRNN) architecture was also proposed for this task, which considered only the sequential relationship within the speech utterances. In this paper, we propose a new Hierarchical Attention Model (HAM), which constructs multi-hopped attention mechanism over tree-structured rather than sequential representations for the utterances. Improved comprehension performance robust with respect to ASR errors were obtained.

Bo-Hsiang Tseng, Sheng-Syun Shen, Hung-Yi Lee et al.

Multimedia or spoken content presents more attractive information than plain text content, but it's more difficult to display on a screen and be selected by a user. As a result, accessing large collections of the former is much more difficult and time-consuming than the latter for humans. It's highly attractive to develop a machine which can automatically understand spoken content and summarize the key information for humans to browse over. In this endeavor, we propose a new task of machine comprehension of spoken content. We define the initial goal as the listening comprehension test of TOEFL, a challenging academic English examination for English learners whose native language is not English. We further propose an Attention-based Multi-hop Recurrent Neural Network (AMRNN) architecture for this task, achieving encouraging results in the initial tests. Initial results also have shown that word-level attention is probably more robust than sentence-level attention for this task with ASR errors.

Yu-An Chung, Chao-Chung Wu, Chia-Hao Shen et al.

The vector representations of fixed dimensionality for words (in text) offered by Word2Vec have been shown to be very useful in many application scenarios, in particular due to the semantic information they carry. This paper proposes a parallel version, the Audio Word2Vec. It offers the vector representations of fixed dimensionality for variable-length audio segments. These vector representations are shown to describe the sequential phonetic structures of the audio segments to a good degree, with very attractive real world applications such as query-by-example Spoken Term Detection (STD). In this STD application, the proposed approach significantly outperformed the conventional Dynamic Time Warping (DTW) based approaches at significantly lower computation requirements. We propose unsupervised learning of Audio Word2Vec from audio data without human annotation using Sequence-to-sequence Audoencoder (SA). SA consists of two RNNs equipped with Long Short-Term Memory (LSTM) units: the first RNN (encoder) maps the input audio sequence into a vector representation of fixed dimensionality, and the second RNN (decoder) maps the representation back to the input audio sequence. The two RNNs are jointly trained by minimizing the reconstruction error. Denoising Sequence-to-sequence Autoencoder (DSA) is furthered proposed offering more robust learning.

Cheng-Tao Chung, Cheng-Yu Tsai, Hsiang-Hung Lu et al.

In this work we aim to discover high quality speech features and linguistic units directly from unlabeled speech data in a zero resource scenario. The results are evaluated using the metrics and corpora proposed in the Zero Resource Speech Challenge organized at Interspeech 2015. A Multi-layered Acoustic Tokenizer (MAT) was proposed for automatic discovery of multiple sets of acoustic tokens from the given corpus. Each acoustic token set is specified by a set of hyperparameters that describe the model configuration. These sets of acoustic tokens carry different characteristics fof the given corpus and the language behind, thus can be mutually reinforced. The multiple sets of token labels are then used as the targets of a Multi-target Deep Neural Network (MDNN) trained on low-level acoustic features. Bottleneck features extracted from the MDNN are then used as the feedback input to the MAT and the MDNN itself in the next iteration. We call this iterative deep learning framework the Multi-layered Acoustic Tokenizing Deep Neural Network (MAT-DNN), which generates both high quality speech features for the Track 1 of the Challenge and acoustic tokens for the Track 2 of the Challenge. In addition, we performed extra experiments on the same corpora on the application of query-by-example spoken term detection. The experimental results showed the iterative deep learning framework of MAT-DNN improved the detection performance due to better underlying speech features and acoustic tokens.

Yi-Hsiu Liao, Hung-yi Lee, Lin-shan Lee

In this paper we propose the Structured Deep Neural Network (structured DNN) as a structured and deep learning framework. This approach can learn to find the best structured object (such as a label sequence) given a structured input (such as a vector sequence) by globally considering the mapping relationships between the structures rather than item by item. When automatic speech recognition is viewed as a special case of such a structured learning problem, where we have the acoustic vector sequence as the input and the phoneme label sequence as the output, it becomes possible to comprehensively learn utterance by utterance as a whole, rather than frame by frame. Structured Support Vector Machine (structured SVM) was proposed to perform ASR with structured learning previously, but limited by the linear nature of SVM. Here we propose structured DNN to use nonlinear transformations in multi-layers as a structured and deep learning approach. This approach was shown to beat structured SVM in preliminary experiments on TIMIT.

Cheng-Tao Chung, Wei-Ning Hsu, Cheng-Yi Lee et al.

This paper presents a novel approach for enhancing the multiple sets of acoustic patterns automatically discovered from a given corpus. In a previous work it was proposed that different HMM configurations (number of states per model, number of distinct models) for the acoustic patterns form a two-dimensional space. Multiple sets of acoustic patterns automatically discovered with the HMM configurations properly located on different points over this two-dimensional space were shown to be complementary to one another, jointly capturing the characteristics of the given corpus. By representing the given corpus as sequences of acoustic patterns on different HMM sets, the pattern indices in these sequences can be relabeled considering the context consistency across the different sequences. Good improvements were observed in preliminary experiments of pattern spoken term detection (STD) performed on both TIMIT and Mandarin Broadcast News with such enhanced patterns.

Cheng-Tao Chung, Chun-an Chan, Lin-shan Lee

This paper presents a new approach for unsupervised Spoken Term Detection with spoken queries using multiple sets of acoustic patterns automatically discovered from the target corpus. The different pattern HMM configurations(number of states per model, number of distinct models, number of Gaussians per state)form a three-dimensional model granularity space. Different sets of acoustic patterns automatically discovered on different points properly distributed over this three-dimensional space are complementary to one another, thus can jointly capture the characteristics of the spoken terms. By representing the spoken content and spoken query as sequences of acoustic patterns, a series of approaches for matching the pattern index sequences while considering the signal variations are developed. In this way, not only the on-line computation load can be reduced, but the signal distributions caused by different speakers and acoustic conditions can be reasonably taken care of. The results indicate that this approach significantly outperformed the unsupervised feature-based DTW baseline by 16.16\% in mean average precision on the TIMIT corpus.

Cheng-Tao Chung, Chun-an Chan, Lin-shan Lee

Techniques for unsupervised discovery of acoustic patterns are getting increasingly attractive, because huge quantities of speech data are becoming available but manual annotations remain hard to acquire. In this paper, we propose an approach for unsupervised discovery of linguistic structure for the target spoken language given raw speech data. This linguistic structure includes two-level (subword-like and word-like) acoustic patterns, the lexicon of word-like patterns in terms of subword-like patterns and the N-gram language model based on word-like patterns. All patterns, models, and parameters can be automatically learned from the unlabelled speech corpus. This is achieved by an initialization step followed by three cascaded stages for acoustic, linguistic, and lexical iterative optimization. The lexicon of word-like patterns defines allowed consecutive sequence of HMMs for subword-like patterns. In each iteration, model training and decoding produces updated labels from which the lexicon and HMMs can be further updated. In this way, model parameters and decoded labels are respectively optimized in each iteration, and the knowledge about the linguistic structure is learned gradually layer after layer. The proposed approach was tested in preliminary experiments on a corpus of Mandarin broadcast news, including a task of spoken term detection with performance compared to a parallel test using models trained in a supervised way. Results show that the proposed system not only yields reasonable performance on its own, but is also complimentary to existing large vocabulary ASR systems.

Cheng-Tao Chung, Cheng-Yu Tsai, Hsiang-Hung Lu et al.

This paper summarizes the work done by the authors for the Zero Resource Speech Challenge organized in the technical program of Interspeech 2015. The goal of the challenge is to discover linguistic units directly from unlabeled speech data. The Multi-layered Acoustic Tokenizer (MAT) proposed in this work automatically discovers multiple sets of acoustic tokens from the given corpus. Each acoustic token set is specified by a set of hyperparameters that describe the model configuration. These sets of acoustic tokens carry different characteristics of the given corpus and the language behind thus can be mutually reinforced. The multiple sets of token labels are then used as the targets of a Multi-target DNN (MDNN) trained on low-level acoustic features. Bottleneck features extracted from the MDNN are used as feedback for the MAT and the MDNN itself. We call this iterative system the Multi-layered Acoustic Tokenizing Deep Neural Network (MAT-DNN) which generates high quality features for track 1 of the challenge and acoustic tokens for track 2 of the challenge.

Bo-Hsiang Tseng, Hung-Yi Lee, Lin-Shan Lee

With the popularity of mobile devices, personalized speech recognizer becomes more realizable today and highly attractive. Each mobile device is primarily used by a single user, so it's possible to have a personalized recognizer well matching to the characteristics of individual user. Although acoustic model personalization has been investigated for decades, much less work have been reported on personalizing language model, probably because of the difficulties in collecting enough personalized corpora. Previous work used the corpora collected from social networks to solve the problem, but constructing a personalized model for each user is troublesome. In this paper, we propose a universal recurrent neural network language model with user characteristic features, so all users share the same model, except each with different user characteristic features. These user characteristic features can be obtained by crowdsouring over social networks, which include huge quantity of texts posted by users with known friend relationships, who may share some subject topics and wording patterns. The preliminary experiments on Facebook corpus showed that this proposed approach not only drastically reduced the model perplexity, but offered very good improvement in recognition accuracy in n-best rescoring tests. This approach also mitigated the data sparseness problem for personalized language models.

Yi-Hsiu Liao, Hung-Yi Lee, Lin-shan Lee

In this paper we propose the Structured Deep Neural Network (Structured DNN) as a structured and deep learning algorithm, learning to find the best structured object (such as a label sequence) given a structured input (such as a vector sequence) by globally considering the mapping relationships between the structure rather than item by item. When automatic speech recognition is viewed as a special case of such a structured learning problem, where we have the acoustic vector sequence as the input and the phoneme label sequence as the output, it becomes possible to comprehensively learned utterance by utterance as a whole, rather than frame by frame. Structured Support Vector Machine (structured SVM) was proposed to perform ASR with structured learning previously, but limited by the linear nature of SVM. Here we propose structured DNN to use nonlinear transformations in multi-layers as a structured and deep learning algorithm. It was shown to beat structured SVM in preliminary experiments on TIMIT.