arXiv subjects

Hermann Ney

Publications and source records attributed to Hermann Ney.

At least 91 records · Page 5Linked to original sources

arXiv2021

Context-Dependent Acoustic Modeling without Explicit Phone Clustering

Phoneme-based acoustic modeling of large vocabulary automatic speech recognition takes advantage of phoneme context. The large number of context-dependent (CD) phonemes and their highly varying statistics require tying or smoothing to enable robust training. Usually, classification and regression trees are used for phonetic clustering, which is standard in hidden Markov model (HMM)-based systems. However, this solution introduces a secondary training objective and does not allow for end-to-end training. In this work, we address a direct phonetic context modeling for the hybrid deep neural network (DNN)/HMM, that does not build on any phone clustering algorithm for the determination of the HMM state inventory. By performing different decompositions of the joint probability of the center phoneme state and its left and right contexts, we obtain a factorized network consisting of different components, trained jointly. Moreover, the representation of the phonetic context for the network relies on phoneme embeddings. The recognition accuracy of our proposed models on the Switchboard task is comparable and outperforms slightly the hybrid model using the standard state-tying decision trees.

eess.AS↗

arXiv2021

A study of latent monotonic attention variants

End-to-end models reach state-of-the-art performance for speech recognition, but global soft attention is not monotonic, which might lead to convergence problems, to instability, to bad generalisation, cannot be used for online streaming, and is also inefficient in calculation. Monotonicity can potentially fix all of this. There are several ad-hoc solutions or heuristics to introduce monotonicity, but a principled introduction is rarely found in literature so far. In this paper, we present a mathematically clean solution to introduce monotonicity, by introducing a new latent variable which represents the audio position or segment boundaries. We compare several monotonic latent models to our global soft attention baseline such as a hard attention model, a local windowed soft attention model, and a segmental soft attention model. We can show that our monotonic models perform as good as the global soft attention model. We perform our experiments on Switchboard 300h. We carefully outline the details of our training and release our code and configs.

cs.CL↗

arXiv2021

Efficient Retrieval Augmented Generation from Unstructured Knowledge for Task-Oriented Dialog

This paper summarizes our work on the first track of the ninth Dialog System Technology Challenge (DSTC 9), "Beyond Domain APIs: Task-oriented Conversational Modeling with Unstructured Knowledge Access". The goal of the task is to generate responses to user turns in a task-oriented dialog that require knowledge from unstructured documents. The task is divided into three subtasks: detection, selection and generation. In order to be compute efficient, we formulate the selection problem in terms of hierarchical classification steps. We achieve our best results with this model. Alternatively, we employ siamese sequence embedding models, referred to as Dense Knowledge Retrieval, to retrieve relevant documents. This method further reduces the computation time by a factor of more than 100x at the cost of degradation in R@1 of 5-6% compared to the first model. Then for either approach, we use Retrieval Augmented Generation to generate responses based on multiple selected snippets and we show how the method can be used to fine-tune trained embeddings.

cs.CL↗

arXiv2020

Two-Way Neural Machine Translation: A Proof of Concept for Bidirectional Translation Modeling using a Two-Dimensional Grid

Neural translation models have proven to be effective in capturing sufficient information from a source sentence and generating a high-quality target sentence. However, it is not easy to get the best effect for bidirectional translation, i.e., both source-to-target and target-to-source translation using a single model. If we exclude some pioneering attempts, such as multilingual systems, all other bidirectional translation approaches are required to train two individual models. This paper proposes to build a single end-to-end bidirectional translation model using a two-dimensional grid, where the left-to-right decoding generates source-to-target, and the bottom-to-up decoding creates target-to-source output. Instead of training two models independently, our approach encourages a single network to jointly learn to translate in both directions. Experiments on the WMT 2018 German$\leftrightarrow$English and Turkish$\leftrightarrow$English translation tasks show that the proposed model is capable of generating a good translation quality and has sufficient potential to direct the research.

cs.CL↗

arXiv2020

Tight Integrated End-to-End Training for Cascaded Speech Translation

A cascaded speech translation model relies on discrete and non-differentiable transcription, which provides a supervision signal from the source side and helps the transformation between source speech and target text. Such modeling suffers from error propagation between ASR and MT models. Direct speech translation is an alternative method to avoid error propagation; however, its performance is often behind the cascade system. To use an intermediate representation and preserve the end-to-end trainability, previous studies have proposed using two-stage models by passing the hidden vectors of the recognizer into the decoder of the MT model and ignoring the MT encoder. This work explores the feasibility of collapsing the entire cascade components into a single end-to-end trainable model by optimizing all parameters of ASR and MT models jointly without ignoring any learned parameters. It is a tightly integrated method that passes renormalized source word posterior distributions as a soft decision instead of one-hot vectors and enables backpropagation. Therefore, it provides both transcriptions and translations and achieves strong consistency between them. Our experiments on four tasks with different data scenarios show that the model outperforms cascade models up to 1.8% in BLEU and 2.0% in TER and is superior compared to direct models.

cs.CL↗

arXiv2020

A New Training Pipeline for an Improved Neural Transducer

The RNN transducer is a promising end-to-end model candidate. We compare the original training criterion with the full marginalization over all alignments, to the commonly used maximum approximation, which simplifies, improves and speeds up our training. We also generalize from the original neural network model and study more powerful models, made possible due to the maximum approximation. We further generalize the output label topology to cover RNN-T, RNA and CTC. We perform several studies among all these aspects, including a study on the effect of external alignments. We find that the transducer model generalizes much better on longer sequences than the attention model. Our final transducer model outperforms our attention model on Switchboard 300h by over 6% relative WER.

eess.AS↗

arXiv2020

Diving Deep into Context-Aware Neural Machine Translation

Context-aware neural machine translation (NMT) is a promising direction to improve the translation quality by making use of the additional context, e.g., document-level translation, or having meta-information. Although there exist various architectures and analyses, the effectiveness of different context-aware NMT models is not well explored yet. This paper analyzes the performance of document-level NMT models on four diverse domains with a varied amount of parallel document-level bilingual data. We conduct a comprehensive set of experiments to investigate the impact of document-level NMT. We find that there is no single best approach to document-level NMT, but rather that different architectures come out on top on different tasks. Looking at task-specific problems, such as pronoun resolution or headline translation, we find improvements in the context-aware systems, even in cases where the corpus-level metrics like BLEU show no significant improvement. We also show that document-level back-translation significantly helps to compensate for the lack of document-level bi-texts.

cs.CL↗

arXiv2020

Robust Beam Search for Encoder-Decoder Attention Based Speech Recognition without Length Bias

As one popular modeling approach for end-to-end speech recognition, attention-based encoder-decoder models are known to suffer the length bias and corresponding beam problem. Different approaches have been applied in simple beam search to ease the problem, most of which are heuristic-based and require considerable tuning. We show that heuristics are not proper modeling refinement, which results in severe performance degradation with largely increased beam sizes. We propose a novel beam search derived from reinterpreting the sequence posterior with an explicit length modeling. By applying the reinterpreted probability together with beam pruning, the obtained final probability leads to a robust model modification, which allows reliable comparison among output sequences of different lengths. Experimental verification on the LibriSpeech corpus shows that the proposed approach solves the length bias problem without heuristics or additional tuning effort. It provides robust decision making and consistently good performance under both small and very large beam sizes. Compared with the best results of the heuristic baseline, the proposed approach achieves the same WER on the 'clean' sets and 4% relative improvement on the 'other' sets. We also show that it is more efficient with the additional derived early stopping criterion.

eess.AS↗

arXiv2020

Successfully Applying the Stabilized Lottery Ticket Hypothesis to the Transformer Architecture

Sparse models require less memory for storage and enable a faster inference by reducing the necessary number of FLOPs. This is relevant both for time-critical and on-device computations using neural networks. The stabilized lottery ticket hypothesis states that networks can be pruned after none or few training iterations, using a mask computed based on the unpruned converged model. On the transformer architecture and the WMT 2014 English-to-German and English-to-French tasks, we show that stabilized lottery ticket pruning performs similar to magnitude pruning for sparsity levels of up to 85%, and propose a new combination of pruning techniques that outperforms all other techniques for even higher levels of sparsity. Furthermore, we confirm that the parameter's initial sign and not its specific value is the primary factor for successful training, and show that magnitude pruning could be used to find winning lottery tickets.

cs.LG↗

arXiv2020

Early Stage LM Integration Using Local and Global Log-Linear Combination

Sequence-to-sequence models with an implicit alignment mechanism (e.g. attention) are closing the performance gap towards traditional hybrid hidden Markov models (HMM) for the task of automatic speech recognition. One important factor to improve word error rate in both cases is the use of an external language model (LM) trained on large text-only corpora. Language model integration is straightforward with the clear separation of acoustic model and language model in classical HMM-based modeling. In contrast, multiple integration schemes have been proposed for attention models. In this work, we present a novel method for language model integration into implicit-alignment based sequence-to-sequence models. Log-linear model combination of acoustic and language model is performed with a per-token renormalization. This allows us to compute the full normalization term efficiently both in training and in testing. This is compared to a global renormalization scheme which is equivalent to applying shallow fusion in training. The proposed methods show good improvements over standard model combination (shallow fusion) on our state-of-the-art Librispeech system. Furthermore, the improvements are persistent even if the LM is exchanged for a more powerful one after training.

eess.AS↗

arXiv2020

Fine-Grained Visual Classification with Efficient End-to-end Localization

The term fine-grained visual classification (FGVC) refers to classification tasks where the classes are very similar and the classification model needs to be able to find subtle differences to make the correct prediction. State-of-the-art approaches often include a localization step designed to help a classification network by localizing the relevant parts of the input images. However, this usually requires multiple iterations or passes through a full classification network or complex training schedules. In this work we present an efficient localization module that can be fused with a classification network in an end-to-end setup. On the one hand the module is trained by the gradient flowing back from the classification network. On the other hand, two self-supervised loss functions are introduced to increase the localization accuracy. We evaluate the new model on the three benchmark datasets CUB200-2011, Stanford Cars and FGVC-Aircraft and are able to achieve competitive recognition performance.

cs.CV↗

arXiv2020

When and Why is Unsupervised Neural Machine Translation Useless?

This paper studies the practicality of the current state-of-the-art unsupervised methods in neural machine translation (NMT). In ten translation tasks with various data settings, we analyze the conditions under which the unsupervised methods fail to produce reasonable translations. We show that their performance is severely affected by linguistic dissimilarity and domain mismatch between source and target monolingual data. Such conditions are common for low-resource language pairs, where unsupervised learning works poorly. In all of our experiments, supervised and semi-supervised baselines with 50k-sentence bilingual data outperform the best unsupervised results. Our analyses pinpoint the limits of the current unsupervised NMT and also suggest immediate research directions.

cs.CL↗

arXiv2020

The RWTH ASR System for TED-LIUM Release 2: Improving Hybrid HMM with SpecAugment

We present a complete training pipeline to build a state-of-the-art hybrid HMM-based ASR system on the 2nd release of the TED-LIUM corpus. Data augmentation using SpecAugment is successfully applied to improve performance on top of our best SAT model using i-vectors. By investigating the effect of different maskings, we achieve improvements from SpecAugment on hybrid HMM models without increasing model size and training time. A subsequent sMBR training is applied to fine-tune the final acoustic model, and both LSTM and Transformer language models are trained and evaluated. Our best system achieves a 5.6% WER on the test set, which outperforms the previous state-of-the-art by 27% relative.

eess.AS↗

arXiv2020

Full-Sum Decoding for Hybrid HMM based Speech Recognition using LSTM Language Model

In hybrid HMM based speech recognition, LSTM language models have been widely applied and achieved large improvements. The theoretical capability of modeling any unlimited context suggests that no recombination should be applied in decoding. This motivates to reconsider full summation over the HMM-state sequences instead of Viterbi approximation in decoding. We explore the potential gain from more accurate probabilities in terms of decision making and apply the full-sum decoding with a modified prefix-tree search framework. The proposed full-sum decoding is evaluated on both Switchboard and Librispeech corpora. Different models using CE and sMBR training criteria are used. Additionally, both MAP and confusion network decoding as approximated variants of general Bayes decision rule are evaluated. Consistent improvements over strong baselines are achieved in almost all cases without extra cost. We also discuss tuning effort, efficiency and some limitations of full-sum decoding.

eess.AS↗

arXiv2020

Generating Synthetic Audio Data for Attention-Based Speech Recognition Systems

Recent advances in text-to-speech (TTS) led to the development of flexible multi-speaker end-to-end TTS systems. We extend state-of-the-art attention-based automatic speech recognition (ASR) systems with synthetic audio generated by a TTS system trained only on the ASR corpora itself. ASR and TTS systems are built separately to show that text-only data can be used to enhance existing end-to-end ASR systems without the necessity of parameter or architecture changes. We compare our method with language model integration of the same text data and with simple data augmentation methods like SpecAugment and show that performance improvements are mostly independent. We achieve improvements of up to 33% relative in word-error-rate (WER) over a strong baseline with data-augmentation in a low-resource environment (LibriSpeech-100h), closing the gap to a comparable oracle experiment by more than 50\%. We also show improvements of up to 5% relative WER over our most recent ASR baseline on LibriSpeech-960h.

cs.CL↗

arXiv2019

A Comparative Study on End-to-end Speech to Text Translation

Recent advances in deep learning show that end-to-end speech to text translation model is a promising approach to direct the speech translation field. In this work, we provide an overview of different end-to-end architectures, as well as the usage of an auxiliary connectionist temporal classification (CTC) loss for better convergence. We also investigate on pre-training variants such as initializing different components of a model using pre-trained models, and their impact on the final performance, which gives boosts up to 4% in BLEU and 5% in TER. Our experiments are performed on 270h IWSLT TED-talks En->De, and 100h LibriSpeech Audiobooks En->Fr. We also show improvements over the current end-to-end state-of-the-art systems on both tasks.

cs.CL↗

arXiv2019

On Using SpecAugment for End-to-End Speech Translation

This work investigates a simple data augmentation technique, SpecAugment, for end-to-end speech translation. SpecAugment is a low-cost implementation method applied directly to the audio input features and it consists of masking blocks of frequency channels, and/or time steps. We apply SpecAugment on end-to-end speech translation tasks and achieve up to +2.2\% \BLEU on LibriSpeech Audiobooks En->Fr and +1.2% on IWSLT TED-talks En->De by alleviating overfitting to some extent. We also examine the effectiveness of the method in a variety of data scenarios and show that the method also leads to significant improvements in various data conditions irrespective of the amount of training data.

cs.CL↗

arXiv2019

On using 2D sequence-to-sequence models for speech recognition

Attention-based sequence-to-sequence models have shown promising results in automatic speech recognition. Using these architectures, one-dimensional input and output sequences are related by an attention approach, thereby replacing more explicit alignment processes, like in classical HMM-based modeling. In contrast, here we apply a novel two-dimensional long short-term memory (2DLSTM) architecture to directly model the input/output relation between audio/feature vector sequences and word sequences. The proposed model is an alternative model such that instead of using any type of attention components, we apply a 2DLSTM layer to assimilate the context from both input observations and output transcriptions. The experimental evaluation on the Switchboard 300h automatic speech recognition task shows word error rates for the 2DLSTM model that are competitive to end-to-end attention-based model.

cs.CL↗