Abstract

In this study, we propose Mel-weighted single frequency filtering (SFF) spectrograms for dialect identification. The spectrum derived using SFF has high spectral resolution for harmonics and resonances while simultaneously maintaining good time-resolution of some speech excitation features such as impulse-like events. The SFF spectrum can represent speech characteristics such as burst time and glottal closure instances better than the short-time Fourier transform (STFT) spectrum. Our hypothesis is that these intricate representations in the SFF spectrum should help in distinguishing dialects. Therefore, we built a dialect identification system which uses an unsupervised, bottleneck feature representation of the Mel-weighted SFF spectrogram (Mel-SFF spectrogram) with sequence-to-sequence deep autoen- coders. The language invariance of the proposed system was evaluated using two datasets: the UT-Podcast database (English) and the STYRIALECT database (German). The proposed representations gave a relative improvement of 9.47% and 4.69% in unweighted average recall (UAR) compared to the best baseline method on the development and test datasets, respectively, of the UT-Podcast database. The proposed representations also gave a comparable performance to the best baseline method for the STYRIALECT database. In addition, the fusion of the autoencoder bottleneck features computed from the Mel-SFF and Mel-STFT spectrograms improved the overall performance indicating complementary information between these features. By further analyzing the performance of the proposed representation with different utterance lengths using the UT-Podcast database, we observed that the proposed representation performed better on short utterances. The improved performance given by the Mel-weighted SFF spectrogram for recognizing dialects in both databases supports our hypothesis

Abstract

Block chain technology provides a decentralized and secure platform for executing transactions. Smart contracts in Ethereum have been proposed as the mechanism to automate legal contracts securely without the involvement of third parties. Yet, there are still several issues to be resolved especially regarding the updating of smart contracts in blockchain as well as the use of blockchain as part of a legal smart contracts system. In this work we propose a methodology and an architecture for building and deploying legal contracts in the blockchain. As the blockchain is immutable, we cannot update the code of the smart legal contracts, but in real life applications updating of contracts is a requirement that cannot be ignored. In this paper we address the problem of contract update by introducing a new versioning system that keeps track of the changes and links the different versions using a linked list. Moreover, we propose a system architecture where the user interface, the application logic and the blockchain are smoothly integrated in a manner that each part of the system contributes for producing a flexible and transparent execution. We show the applicability of our approach by implementing a system for the case of a rental agreement.

Abstract

In production of voiced speech, epochs or glottal closure instants (GCIs) refer to the instants of significant excitation of the vocal tract. Extraction of GCIs is used as a pre-processing stage in many areas of speech technology, such as in prosody modification, speech syn- thesis and voice source analysis. In the past decades, several GCI detection algorithms have been developed and most of them provide excellent results for speech signals produced using modal (normal) type of phonation. There are, however, no studies comparing multi- ple state-of-the-art GCI detection methods in emotional speech. In this paper, we compare six GCI detection algorithms using emo- tional speech and known evaluation metrics. We use the Berlin EMO-DB acted emotional speech database which contains seven emotions and simultaneous electroglottography (EGG) recordings as ground truth. The results show that all six GCI detection algo- rithms give best performance in processing speech of neutral emo- tion and that the performance degrade particularly in emotions of high arousal (anger and joy). To improve the performance of GCI detection in emotional speech, the study underlines the importance of local average pitch period estimates.

Abstract

Multi user massive multiple input multiple output (MU-m-MIMO) has emerged as a viable technology for scaling up existing communication systems, and in serving increasing number of users for the next-generation communication systems. Several signal processing algorithms exist for mitigating the performance-limiting artefacts encountered in MU-m-MIMO systems (like inter-symbol interference, inter-channel interfer- ence, and device nonlinearities), among which, reproducing kernel Hilbert space (RKHS) based approaches have emerged to provide effective solutions. However, most of the existing RKHS based detectors for MU-m-MIMO are dictionary-based, which makes it difficult to gauge the memory requirements beforehand, and are prone to error in the presence of noisy observations. Hence, to reduce the computational complexity, a Random Fourier Features (RFF) based parallel detection algorithm is proposed for MU-m-MIMO, that uses decomposed blocks of high dimensional observations, and makes the proposed detector scalable for parallel computation using modern multi- core compute-units at the receivers (which is possible today due to advances in computing). Further, the RFF based explicit feature map to RKHS alleviates the requirement of a dictionary, and facilitates ease of practical implementation. Simulations are performed over realistic MU-m-MIMO systems, which indicates that the proposed approach delivers an acceptable uncoded BER performance, whilst maintaining a finite implementation budget, which makes the proposed approach attractive for implementation. Lastly, the error-rate analysis of the proposed detector is performed, and validated through simulations

Abstract

Most of the applications in speech use mel-frequency spectral coefficients (MFSC) as features as they match the human perceptual mechanism, where the emphasis is given to vocal tract characteristics. But in accent classification, mel-scale distribution of filters may not always be the best representations, e.g., pitch accented languages where the emphasis should be on vocal source information too. Motivated by this, we use end-to-end classification of accents directly from waveforms which will reduce the effort of designing features specific to each corpus. The convolution neural network (CNN) model architecture is designed in such a way that the initial layers exhibit similar operation as in MFSC by initializing the weights using time approximate of MFSC. The entire network along with initial layers is trained to learn accent classification. We observed that learning directly from waveform improved the performance of accent classification when compared to CNN trained on hand-engineered features by 10.94% UAR on the test dataset of common voice corpus. Analyzing the filters after learning, we observed changes in distribution and bandwidths of center frequencies. We further observed the importance of appropriately initializing CNN filters

Abstract

Various representations have been developed for acoustic scene classifications (ASC) task using spectral information. However, there is a wide gap in dealing with acoustic scene representations. In this paper, we propose to use a single frequency filtering (SFF) approach, which provides good temporal and spectral resolution at each instant. Single-frequency filtering cepstral coefficients (SFFCC) with deep neural network (DNN) model as the classifier is used for the experimentation on DCASE 2019 and DCASE 2018 Task 1, development data of subtasks A and B. From the conducted experiments on the development datasets, the usage of the SFFCC features significantly improved ASC performance. This approach has got 35 th team rank out of 46 submissions to the corresponding DCASE 2019 Task 1A challenge with a 52.6% classification accuracy on the evaluation dataset. Also, the effect of raw waveforms taken as features for ASC using DNNs was observed.

Abstract

The past few years have seen several works exploring learning economic solutions from data; these include optimal auction design, function optimization, stable payoffs in cooperative games and more. In this work, we provide a unified learning-theoretic methodology for modeling such problems, and establish tools for determining whether a given solution concept can be efficiently learned from data. Our learning theoretic framework generalizes a notion of function space dimension — the graph dimension — adapting it to the solution concept learning domain. We identify sufficient conditions for efficient solution learnability, and show that results in existing works can be immediately derived using our methodology. Finally, we apply our methods in other economic domains, yielding learning variants of competitive equilibria and Condorcet winners

Abstract

We investigate a growing body of work that seeks to improve recommender systems through the use of review text. Generally, these papers argue that since reviews ‘explain’ users’ opinions, they ought to be useful to infer the underlying dimensions that predict ratings or purchases. Schemes to incorporate reviews range from simple regularizers to neural network approaches. Our initial findings reveal several discrepancies in reported results, partly due to (e.g.) copying results across papers despite changes in experimental settings or data pre-processing. First, we attempt a comprehensive analysis to resolve these ambiguities. Further investigation calls for discussion on a much larger problem about the “importance" of user reviews for recommendation. Through a wide range of experiments, we observe several cases where state-of-the-art methods fail to outperform existing baselines, especially as we deviate from a few narrowly-defined settings where reviews are useful. We conclude by providing hypotheses for our observations, that seek to characterize under what conditions reviews are likely to be helpful. Through this work, we aim to evaluate the direction in which the field is progressing and encourage robust empirical evaluation.

Abstract

We investigate a growing body of work that seeks to improve recommender systems through the use of review text. Generally, these papers argue that since reviews ‘explain’ users’ opinions, they ought to be useful to infer the underlying dimensions that predict ratings or purchases. Schemes to incorporate reviews range from simple regularizers to neural network approaches. Our initial findings reveal several discrepancies in reported results, partly due to (e.g.) copying results across papers despite changes in experimental settings or data pre-processing. First, we attempt a comprehensive analysis to resolve these ambiguities. Further investigation calls for discussion on a much larger problem about the “importance" of user reviews for recommendation. Through a wide range of experiments, we observe several cases where state-of-the-art methods fail to outperform existing baselines, especially as we deviate from a few narrowly-defined settings where reviews are useful. We conclude by providing hypotheses for our observations, that seek to characterize under what conditions reviews are likely to be helpful. Through this work, we aim to evaluate the direction in which the field is progressing and encourage robust empirical evaluation.

Abstract

Learning effective contextual-bandit policies from past actions of a deployed system is highly desirable in many settings (e.g. voice assistants, recommendation, search), since it enables the reuse of large amounts of log data. State-of-the-art methods for such offpolicy learning, however, are based on inverse propensity score (IPS) weighting. A key theoretical requirement of IPS weighting is that the policy that logged the data has "full support", which typically translates into requiring non-zero probability for any action in any context. Unfortunately, many real-world systems produce support deficient data, especially when the action space is large, and we show how existing methods can fail catastrophically. To overcome this gap between theory and applications, we identify three approaches that provide various guarantees for IPS-based learning despite the inherent limitations of support-deficient data: restricting the action space, reward extrapolation, and restricting the policy space. We systematically analyze the statistical and computational properties of these three approaches, and we empirically evaluate their effectiveness. In addition to providing the first systematic analysis of support-deficiency in contextual-bandit learning, we conclude with recommendations that provide practical guidance.