For children, the system trained on a large corpus of adult speakers performed worse than a system trained on a much smaller corpus of children’s speech. This is due to the acoustic mismatch between training and testing data. To capture more acoustic variability we trained a shared system with mixed data from adults and children. The shared system yields the best EER for children with no degradation for adults. Thus, the single system trained with mixed data is applicable for speaker verification for both adults and children.

In this paper, we propose spectral modification by sharpening formants and by reducing the spectral tilt to recognize children’s speech by automatic speech recognition (ASR) systems developed using adult speech. In this type of mismatched condition, the ASR performance is degraded due to the acoustic and linguistic mismatch in the attributes between children and adult speakers. The proposed method is used to improve the speech intelligibility to enhance the children’s speech recognition using an acoustic model trained on adult speech. In the experiments, WSJCAM0 and PFSTAR are used as databases for adults’ and children’s speech, respectively. The proposed technique gives a significant improvement in the context of the DNN-HMM-based ASR. Furthermore, we validate the robustness of the technique by showing that it performs well also in mismatched noise conditions.

Forced alignment is an effective process to speed up linguistic research. However, most forced aligners are language-dependent, and under-resourced languages rarely have enough resources to train an acoustic model for an aligner. We present a new Finnish grapheme-based forced aligner and demonstrate its performance by aligning multiple Uralic languages and English as an unrelated language. We show that even a simple non-expert created grapheme-to-phoneme mapping can result in useful word alignments.

We propose a simple and efficient framework to learn syntactic embeddings based on information derived from constituency parse trees. Using biased random walk methods, our embeddings not only encode syntactic information about words, but they also capture contextual information. We also propose a method to train the embeddings on multiple constituency parse trees to ensure the encoding of global syntactic representation. Quantitative evaluation of the embeddings show a competitive performance on POS tagging task when compared to other types of embeddings, and qualitative evaluation reveals interesting facts about the syntactic typology learned by these embeddings.

Character-based Neural Network Language Models (NNLM) have the advantage of smaller vocabulary and thus faster training times in comparison to NNLMs based on multi-character units. However, in low-resource scenarios, both the character and multi-character NNLMs suffer from data sparsity. In such scenarios, cross-lingual transfer has improved multi-character NNLM performance by allowing information transfer from a source to the target language. In the same vein, we propose to use cross-lingual transfer for character NNLMs applied to low-resource Automatic Speech Recognition (ASR). However, applying cross-lingual transfer to character NNLMs is not as straightforward. We observe that relatedness of the source language plays an important role in cross-lingual pretraining of character NNLMs. We evaluate this aspect on ASR tasks for two target languages: Finnish (with English and Estonian as source) and Swedish (with Danish, Norwegian, and English as source). Prior work has observed no difference between using the related or unrelated language for multi-character NNLMs. We, however, show that for character-based NNLMs, only pretraining with a related language improves the ASR performance, and using an unrelated language may deteriorate it. We also observe that the benefits are larger when there is much lesser target data than source data.

Crowdsourcing is the go-to solution for data collection and annotation in the context of NLP tasks. Nevertheless, crowdsourced data is noisy by nature; the source is often unknown and additional validation work is performed to guarantee the dataset’s quality. In this article, we compare two crowdsourcing sources on a dialogue paraphrasing task revolving around a chatbot service. We observe that workers hired on crowdsourcing platforms produce lexically poorer and less diverse rewrites than service users engaged voluntarily. Notably enough, on dialogue clarity and optimality, the two paraphrase sources’ human-perceived quality does not differ significantly. Furthermore, for the chatbot service, the combined crowdsourced data is enough to train a transformer-based Natural Language Generation (NLG) system. To enable similar services, we also release tools for collecting data and training the dialogue-act-based transformer-based NLG module.

Data-driven segmentation of words into subword units has been used in various natural language processing applications such as automatic speech recognition and statistical machine translation for almost 20 years. Recently it has became more widely adopted, as models based on deep neural networks often benefit from subword units even for morphologically simpler languages. In this paper, we discuss and compare training algorithms for a unigram subword model, based on the Expectation Maximization algorithm and lexicon pruning. Using English, Finnish, North Sami, and Turkish data sets, we show that this approach is able to find better solutions to the optimization problem defined by the Morfessor Baseline model than its original recursive training algorithm. The improved optimization also leads to higher morphological segmentation accuracy when compared to a linguistic gold standard. We publish implementations of the new algorithms in the widely-used Morfessor software package.

Participating in conversations can be difficult for people with hearing loss, especially in acoustically challenging environments. We studied the preferences the hearing impaired have for a personal conversation assistant based on automatic speech recognition (ASR) technology. We created two prototypes which were evaluated by hearing impaired test users. This paper qualitatively compares the two based on the feedback obtained from the tests. The first prototype was a proof-of-concept system running real-time ASR on a laptop. The second prototype was developed for a mobile device with the recognizer running on a separate server. In the mobile device, augmented reality (AR) was used to help the hearing impaired observe gestures and lip movements of the speaker simultaneously with the transcriptions. Several testers found the systems useful enough to use in their daily lives, with majority preferring the mobile AR version. The biggest concern of the testers was the accuracy of the transcriptions and the lack of speaker identification.

This article describes the Aalto University entry to the WMT18 News Translation Shared Task. We participate in the multilingual subtrack with a system trained under the constrained condition to translate from English to both Finnish and Estonian. The system is based on the Transformer model. We focus on improving the consistency of morphological segmentation for words that are similar orthographically, semantically, and distributionally; such words include etymological cognates, loan words, and proper names. For this, we introduce Cognate Morfessor, a multilingual variant of the Morfessor method. We show that our approach improves the translation quality particularly for Estonian, which has less resources for training the translation model.

This paper describes the MeMAD project entry to the WMT Multimodal Machine Translation Shared Task. We propose adapting the Transformer neural machine translation (NMT) architecture to a multi-modal setting. In this paper, we also describe the preliminary experiments with text-only translation systems leading us up to this choice. We have the top scoring system for both English-to-German and English-to-French, according to the automatic metrics for flickr18. Our experiments show that the effect of the visual features in our system is small. Our largest gains come from the quality of the underlying text-only NMT system. We find that appropriate use of additional data is effective.

This paper describes the MeMAD project entry to the IWSLT Speech Translation Shared Task, addressing the translation of English audio into German text. Between the pipeline and end-to-end model tracks, we participated only in the former, with three contrastive systems. We tried also the latter, but were not able to finish our end-to-end model in time. All of our systems start by transcribing the audio into text through an automatic speech recognition (ASR) model trained on the TED-LIUM English Speech Recognition Corpus (TED-LIUM). Afterwards, we feed the transcripts into English-German text-based neural machine translation (NMT) models. Our systems employ three different translation models trained on separate training sets compiled from the English-German part of the TED Speech Translation Corpus (TED-TRANS) and the OPENSUBTITLES2018 section of the OPUS collection. In this paper, we also describe the experiments leading up to our final systems. Our experiments indicate that using OPENSUBTITLES2018 in training significantly improves translation performance. We also experimented with various preand postprocessing routines for the NMT module, but we did not have much success with these. Our best-scoring system attains a BLEU score of 16.45 on the test set for this year’s task.

String segmentation is an important and recurring problem in natural language processing and other domains. For morphologically rich languages, the amount of different word forms caused by morphological processes like agglutination, compounding and inflection, may be huge and causes problems for traditional word-based language modeling approach. Segmenting text into better modelable units is thus an important part of the modeling task. This work presents methods and a toolkit for learning segmentation models from text. The methods may be applied to lexical unit selection for speech recognition and also other segmentation tasks.

Current ASR and MT systems do not operate on conversational Finnish, because training data for colloquial Finnish has not been available. Although speech recognition performance on literary Finnish is already quite good, those systems have very poor baseline performance in conversational speech. Text data for relevant vocabulary and language models can be collected from the Internet, but web data is very noisy and most of it is not helpful for learning good models. Finnish language is highly agglutinative, and written phonetically. Even phonetic reductions and sandhi are often written down in informal discussions. This increases vocabulary size dramatically and causes word-based selection methods to fail. Our selection method explicitly optimizes the perplexity of a subword language model on the development data, and requires only very limited amount of speech transcripts as development data. The language models have been evaluated for speech recognition using a new data set consisting of generic colloquial Finnish.