Graphemes outperform phonemes in end-to-end models for English Voice-search and multi-dialect tasks.
problem Comparing phoneme-based and grapheme-based sub-word units in end-to-end models.
method Detailed experiments comparing phoneme-based and grapheme-based end-to-end models on large vocabulary English Voice-search and multi-dialect tasks.
result Graphemes outperform phonemes in end-to-end models for English Voice-search and multi-dialect tasks.
Improved online neural transducer model matches non-streaming model performance.
problem Significant performance degradation of online neural transducer models.
method Increased attention window, LAS initialization, stronger language models.
result Improved online neural transducer model matches non-streaming model performance.
MPSA-DenseNet improves accent classification accuracy.
problem Accurate English accent identification.
method Combines multi-task learning and PSA attention mechanism with DenseNet.
result MPSA-DenseNet outperforms other models in accent classification.
Improved speech recognition model reduces error rate from 9.2% to 5.6%.
problem Improving speech recognition accuracy for voice search tasks.
method Structural and optimization improvements to Listen, Attend, and Spell (LAS) model.
result Significant reduction in word error rate (WER) from 9.2% to 5.6% on voice search task.
Improved ASR for English-isiZulu code-switched speech with semi-supervised training.
problem Improving ASR for code-switched speech between English and isiZulu.
method Semi-supervised training using automatic transcription of multilingual speech data.
result Semi-supervised training achieved significant WER reduction in ASR performance.
Paper explores using hand gestures to type English letters.
problem Creating easy-to-remember, non-cumbersome gestures for typing.
method Statistical approach to handle randomness in hand movements.
result Achieved 97.33% accuracy with entire English alphabet.
Improved spoken English intelligibility with computer recognition and feature extraction.
problem Improving spoken English pronunciation and intelligibility.
method Automatic speech recognition using PocketSphinx alignment and feature extraction with SVM classifier probability prediction.
result SVM models achieve 82 percent agreement with human transcriptions, up from 75 percent.
Study examines bias in language models across multiple languages.
problem Assessing bias in language models across different languages.
method Semi-automatically translated data sets into multiple languages, analyzed mono- and multilingual models.
result Notable differences in bias across languages, with Turkish models showing least stereotypes.
The goal of this modern presentation, followed by an English translation from the German, is to make available some parts of Lie's very systematic mathematical thought which deserve to join the contemporary literature, and above all also, to be read.
Evolved Transformer improves on Transformer architecture for language tasks.
problem Improving Transformer architecture for sequence tasks.
method Evolutionary architecture search with warm starting and dynamic resource allocation.
result Evolved Transformer achieves state-of-the-art BLEU scores and reduces parameter count.
In this paper, we attempt to improve Statistical Machine Translation (SMT) systems on a very diverse set of language pairs (in both directions): Czech - English, Vietnamese - English, French - English and German - English. To accomplish this, we performed translation model training, created adaptations of training sett…
Memory-augmented neural networks improve machine translation performance.
problem Improving machine translation accuracy and flexibility.
method Evaluation of Neural Turing Machines and Differentiable Neural Computers for machine translation tasks.
result Memory-augmented neural networks perform similarly to attentional encoders on Vietnamese to English tasks but have lower BLEU scores on Romanian to English tasks.
In this paper, we present Neural Phrase-based Machine Translation (NPMT). Our method explicitly models the phrase structures in output sequences using Sleep-WAke Networks (SWAN), a recently proposed segmentation-based sequence modeling method. To mitigate the monotonic alignment requirement of SWAN, we introduce a new …
Paper shows continuous speech recognition with EEG features, no speech input.
problem Continuous speech recognition with limited vocabulary and noisy/no speech input.
method Connectionist temporal classification (CTC) model, EEG features, new deep learning architecture.
result Continuous speech recognition achieved on limited vocabulary with noisy/no speech input.
This study improves NMT using reinforcement learning, overcoming its instability.
problem Stability issues in reinforcement learning for neural machine translation.
method Systematic study on reinforcement learning factors and a new method for monolingual data.
result Competitive results on WMT17 Chinese-English translation task, setting a state-of-the-art performance.
Proposes a multilingual email segmentation benchmark and model.
problem Lack of multilingual email zoning corpora and models.
method Analysis of existing corpora, development of multilingual benchmark, introduction of OKAPI model.
result OKAPI model achieves state-of-the-art performance in English and generalizes well to unseen languages.
This research explores the effects of various training settings on a Polish to English Statistical Machine Translation system for spoken language. Various elements of the TED, Europarl, and OPUS parallel text corpora were used as the basis for training of language models, for development, tuning and testing of the tran…
We present Listen, Attend and Spell (LAS), a neural network that learns to transcribe speech utterances to characters. Unlike traditional DNN-HMM models, this model learns all the components of a speech recognizer jointly. Our system has two components: a listener and a speller. The listener is a pyramidal recurrent ne…
Hybrid ASR systems can model graphemes effectively using chenones, outperforming traditional methods.
problem Traditional hybrid ASR systems struggle with English's poor grapheme-phoneme correspondence.
method Leveraging tied context-dependent graphemes (chenones) to model graphemes directly.
result Chenone-based systems significantly outperform senone baselines by 4.5% to 11.1% on English datasets.
We show that the predictability of letters in written English texts depends strongly on their position in the word. The first letters are usually the least easy to predict. This agrees with the intuitive notion that words are well defined subunits in written languages, with much weaker correlations across these units t…
Tensor Train layer improves BLEU scores in NMT models.
problem Improving Neural Machine Translation (NMT) models' performance.
method Implemented Tensor Train layer in TensorFlow for NMT training.
result Higher learning rates and more 'rectangular' core dimensions improve BLEU scores.
Continuous speech recognition from brain activity without vocalization.
problem Recognizing silent speech from EEG signals.
method Implemented a CTC ASR model using EEG signals.
result Demonstrated feasibility of EEG for continuous silent speech recognition.
Improved speaker diarization with LSTM and d-vectors.
problem Improving speaker diarization performance.
method Combining LSTM-based d-vector audio embeddings with non-parametric clustering.
result Achieved state-of-the-art diarization error rate of 12.0% on NIST SRE 2000 CALLHOME.
This paper closes the accuracy gap between A2W models and sub-word models using English conversational speech data.
problem Closing the accuracy gap between direct acoustics-to-word models and sub-word models.
method Training an A2W model with orders of magnitude more data, optimizing model initialization, training data order, and regularization.
result Achieved word error rates of 8.8%/13.9% on Hub5-2000 Switchboard/CallHome test sets.
RNN models classify poem meters from plain text with high accuracy.
problem Classifying poem meters from plain text.
method Character-level encoding, RNN models, no feature handcrafting.
result 96.38% accuracy for Arabic poems, 82.31% for English poems.
Dataset for measuring reading levels in India's children.
problem Measuring reading levels in India's vast population.
method Developed ASER dataset with 5,301 subjects in Hindi, Marathi, and English.
result Achieved 86% accuracy in English language classification.
We describe a unified and coherent syntactic framework for supporting a semantically-informed syntactic approach to statistical machine translation. Semantically enriched syntactic tags assigned to the target-language training texts improved translation quality. The resulting system significantly outperformed a linguis…
This research explores the effects of various training settings from Polish to English Statistical Machine Translation system for spoken language. Various elements of the TED parallel text corpora for the IWSLT 2013 evaluation campaign were used as the basis for training of language models, and for development, tuning …
This is the less official, English version of the proof of the fact that every closed atoroidal 3-manifold carries finitely many isotopy classes of tight contact structures.
Comment classification on cookery channels using BERT and traditional models.
problem Volume of multilingual comments, variable lengths, slang, symbols, and abbreviations make comment classification challenging.
method Evaluated traditional machine learning models (Naive Bayes, KNN, SVM, Random Forest, Decision Trees) and BERT-based models (BERT, DISTILBERT, XLM) for multilingual comment classification.
result XLM was the top-performing BERT model with an accuracy of 67.31, while Random Forest with Term Frequency Vectorizer was the best traditional model with 63.59 accuracy.
Project develops lip reading algorithm for limited English.
problem Limited audio information for lip reading.
method Extract lip positions from video frames, classify visemes and phonemes, use HMMs to predict words.
result Algorithm predicts words from lip movements for a subset of English.
Investigates neural TTS systems for Japanese and English.
problem Improving neural TTS systems for high-quality speech synthesis.
method Comparative study of neural sequence-to-sequence TTS vs. DNN pipeline TTS, varying model architecture, parameter size, and language.
result A neural sequence-to-sequence TTS system requires sufficient model parameters and a powerful encoder for high-quality speech synthesis.
Develops a cross-lingual hate speech detection model using pre-trained Transformers.
problem Detecting hate speech in low-resource languages.
method Utilizes frozen Transformer language models and AXEL attention-based classification block for zero-shot and few-shot learning.
result Demonstrates highly competitive results on English and Spanish subsets of the HatEval challenge.
Amobee won 3rd and 1st place in SemEval 2018 sentiment classification tasks.
problem Sentiment classification in multiple languages.
method Training GRU-CNN model with word embeddings and stacking ensembles.
result 3rd and 1st place in valence ordinal classification sub-tasks in English and Spanish.
Transformer++ improves neural machine translation BLEU scores.
problem Handling long-range dependencies in sentences.
method Proposes a new multi-head attention mechanism with context learning.
result Achieves new state-of-the-art BLEU scores on English-to-German and English-to-French translation tasks.
We present a novel method for solving Canonical Correlation Analysis (CCA) in a sparse convex framework using a least squares approach. The presented method focuses on the scenario when one is interested in (or limited to) a primal representation for the first view while having a dual representation for the second view…
This new research explores the effects of various training methods on a Polish to English Statistical Machine Translation system for medical texts. Various elements of the EMEA parallel text corpora from the OPUS project were used as the basis for training of phrase tables and language models and for development, tunin…
Many supervised learning tasks are emerged in dual forms, e.g., English-to-French translation vs. French-to-English translation, speech recognition vs. text to speech, and image classification vs. image generation. Two dual tasks have intrinsic connections with each other due to the probabilistic correlation between th…
BERT model improves cross-lingual document retrieval.
problem Improving cross-lingual information retrieval between English queries and foreign documents.
method Introduced a deep relevance matching model based on BERT, trained with weak supervision using CLIR data.
result Model outperforms competitive baseline approaches in retrieving Lithuanian documents against short English queries.
Flexible method for estimating frequencies in large datasets using sketching.
problem Estimating frequencies in very large datasets efficiently and accurately.
method Data-adaptive conformal inference method based on a smaller sketch of data.
result Proves validity of frequentist confidence intervals under data exchangeability.
Paper identifies and solves a 'scrambled translation' issue in UNMT models.
problem Scrambled translation in UNMT models using word shuffle noise.
method Retraining UNMT models without noise after a pre-defined number of iterations.
result Retraining strategy improves BLEU scores by 10-20% on various language pairs.
Improved neural model predicts gender from tweets.
problem Predicting gender from Twitter text.
method RNN model with attention, LSA-reduced n-gram features.
result Improved model achieves state-of-the-art performance on English tweets.
Wi-Fringe recognizes named gestures using WiFi CSI without training data.
problem Lack of training data for WiFi-based gesture recognition.
method WiFi CSI values paired with activity names for recognition.
result Wi-Fringe detects activities with or without training data.
EBR improves NMT by re-ranking samples drawn from MLE-trained models.
problem Discrepancy between MLE and BLEU score in neural machine translation.
method Train an energy-based model to mimic BLEU score, then use it for re-ranking.
result EBR consistently improves NMT performance by +4 BLEU points on IWSLT'14 German-English.
This paper optimizes sports betting strategies using neural networks and portfolio theory.
problem Optimizing betting strategies in sports gambling.
method Combining neural network models with portfolio optimization, integrating Von Neumann-Morgenstern Expected Utility Theory and the Kelly Criterion.
result Achieved 135.8% relative profit during the English Premier League season.
Automates translating natural language to Verilog for digital design.
problem Manual translation of natural language specifications to Verilog is time-consuming and error-prone.
method Fine-tuned GPT-2 to derive Verilog from English, using a dataset of design tasks.
result GPT-2 achieved 94.8% correct translation across simple and abstract design tasks.
Cricket betting is a multi-billion dollar market. Therefore, there is a strong incentive for models that can predict the outcomes of games and beat the odds provided by bookers. The aim of this study was to investigate to what degree it is possible to predict the outcome of cricket matches. The target competition was t…
This research tackles uncertainty estimation in autoregressive structured prediction tasks.
problem Ensuring safety and robustness of AI systems through accurate uncertainty estimation.
method Develops a unified probabilistic ensemble-based framework for token-level and sequence-level uncertainty estimation.
result Provides baselines for error and out-of-domain detection on translation and speech recognition datasets.