检索结果-内蒙古大学图书馆

Reducing over-smoothness in speech synthesis using Generative Adversarial Networks

学校读者我要写书评

暂无评论

Reducing over-smoothness in speech synthesis using Generativ...

IEEE Region 8 International Conference on Computational Technologies in Electrical and Electronics Engineering, SIBIRCON

作者： Leyuan Sheng Evgeniy N. Pavlovskiy Novosibirsk State University Novosibirsk Russia Stream Data Analytics and Machine Learning laboratory Novosibirsk State University Novosibirsk Russia

Speech synthesis is widely used in many practical applications. In recent years, speech synthesis technology has developed rapidly. However, one of the reasons why synthetic speech is unnatural is that it often has over-smoothness. In order to improve the naturalness of synthetic speech, we first extract the mel-spectrogram of speech and convert it into a real image, then take the over-smooth mel-spectrogram image as input, and use image-to-image translation Generative Adversarial Networks(GANs) framework to generate a more realistic mel-spectrogram. Finally, the results show that this method greatly reduces the over-smoothness of synthesized speech and is more close to the mel-spectrogram of real speech.

关键词： Generative adversarial networks Speech synthesis Spectrogram Generators Hidden Markov models Training

p-value adjustment for monotonous, unbiased, and fast clustering comparison 23

学校读者我要写书评

暂无评论

p-value adjustment for monotonous, unbiased, and fast cluste...

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Kai Klede Thomas Altstidl Dario Zanca Björn Eskofier Machine Learning and Data Analytics (MaD) Lab Friedrich-Alexander Universität Erlangen-Nürnberg Machine Learning and Data Analytics (MaD) Lab Friedrich-Alexander Universität Erlangen-Nürnberg and Translational Digital Health Group Institute of AI for Health Helmholtz Zentrum München

Popular metrics for clustering comparison, like the Adjusted Rand Index and the Adjusted Mutual Information, are type II biased. The Standardized Mutual Information removes this bias but suffers from counterintuitive non-monotonicity and poor computational efficiency. We introduce the p-value adjusted Rand Index (PMI2), the first cluster comparison method that is type II unbiased and provably monotonous. The PMI2 has fast approximations that outperform the Standardized Mutual information. We demonstrate its unbiased clustering selection, approximation quality, and runtime efficiency on synthetic benchmarks. In experiments on image and social network datasets, we show how the PMI2 can help practitioners choose better clustering and community detection algorithms.

关键词：

Thai Conversational Chatbot Classification Using BiLSTM and data Augmentation 1st

学校读者我要写书评

暂无评论

Thai Conversational Chatbot Classification Using BiLSTM and ...

1st International Conference on data Science and Artificial Intelligence, DSAI 2023

作者： Lhasiw, Nunthawat Tanantong, Tanatorn Sanglerdsinlapachai, Nuttapong Thammasat Research Unit in Data Innovation and Artificial Intelligence Department of Computer Science Faculty of Science and Technology Thammasat University Pathum Thani Thailand Strategic Analytics Networks with Machine Learning and AI Research Team National Electronics and Computer Technology Center Pathum Thani Thailand

ISBN: (纸本)9789819979684

Chatbot platforms, e.g., Facebook and Line, have revolutionized human interaction in the digital age. In order to develop an automatic chatbot classification, there are several challenges especially for Thai chat messages. Conversational messages are usually short and ambiguous. Therefore, it is difficult to find a dataset for constructing an effective classification model. To address the limited size of the dataset, data augmentation techniques can be possibly applied. data augmentation involves generating synthetic messages by applying various transformations to existing data samples while preserving their original meaning. In this study, the size and diversity of the dataset is increased by two methods, i.e., text augmentation using word2vec from Thai2Fit and English-Thai machine translation models proposed by VISTEC. Based on the augmented messages, a Deep learning technique, BiLSTM, is used to construct a chatbot classification model. The experimental obtained results demonstrate that data augmentation can help to increase the classification performance. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2023.

关键词： Classification (of information)

CLIP: Cheap Lipschitz Training of Neural Networks 8th

学校读者我要写书评

暂无评论

CLIP: Cheap Lipschitz Training of Neural Networks

8th International Conference on Scale Space and Variational Methods in Computer Vision, SSVM 2021

作者： Bungert, Leon Raab, René Roith, Tim Schwinn, Leo Tenbrinck, Daniel Department Mathematics Friedrich-Alexander University Erlangen-Nürnberg Cauerstraße 11 Erlangen91058 Germany Machine Learning and Data Analytics Lab Friedrich-Alexander University Erlangen-Nürnberg Carl-Thiersch-Straße 2b Erlangen91052 Germany

ISBN: (纸本)9783030755485

Despite the large success of deep neural networks (DNN) in recent years, most neural networks still lack mathematical guarantees in terms of stability. For instance, DNNs are vulnerable to small or even imperceptible input perturbations, so called adversarial examples, that can cause false predictions. This instability can have severe consequences in applications which influence the health and safety of humans, e.g., biomedical imaging or autonomous driving. While bounding the Lipschitz constant of a neural network improves stability, most methods rely on restricting the Lipschitz constants of each layer which gives a poor bound for the actual Lipschitz constant. In this paper we investigate a variational regularization method named CLIP for controlling the Lipschitz constant of a neural network, which can easily be integrated into the training procedure. We mathematically analyze the proposed model, in particular discussing the impact of the chosen regularization parameter on the output of the network. Finally, we numerically evaluate our method on both a nonlinear regression problem and the MNIST and Fashion-MNIST classification databases, and compare our results with a weight regularization approach. © 2021, Springer Nature Switzerland AG.

关键词： Stability

Modeling lens potentials with continuous neural fields in galaxy-scale strong lenses

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Biggio, Luca Vernardos, Georgios Galan, Aymeric Peel, Austin Data Analytics Lab Institute of Machine Learning Department of Computer Science ETHZ Switzerland Observatoire de Sauverny Versoix1290 Switzerland

Strong gravitational lensing is a unique observational tool for studying the dark and luminous mass distribution both within and between galaxies. Given the presence of substructures, current strong lensing observations demand more complex mass models than smooth analytical profiles, such as power-law ellipsoids. In this work, we introduce a continuous neural field to predict the lensing potential at any position throughout the image plane, allowing for a nearly model-independent description of the lensing mass. We apply our method on simulated Hubble Space Telescope imaging data containing different types of perturbations to a smooth mass distribution: a localized dark subhalo, a population of subhalos, and an external shear perturbation. Assuming knowledge of the source surface brightness, we use the continuous neural field to model either the perturbations alone or the full lensing potential. In both cases, the resulting model is able to fit the imaging data, and we are able to accurately recover the properties of both the smooth potential and of the perturbations. Unlike many other deep learning methods, ours explicitly retains lensing physics (i.e., the lens equation) and introduces high flexibility in the model only where required, namely, in the lens potential. Moreover, the neural network does not require pre-training on large sets of labelled data and predicts the potential from the single observed lensing image. Our model is implemented in the fully differentiable lens modeling code Herculens. © 2022, CC BY.

关键词： Gravitation

Towards an IMU-based pen online handwriting recognizer

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Wehbi, Mohamad Hamann, Tim Barth, Jens Kaempf, Peter Zanca, Dario Eskofier, Bjoern Machine Learning and Data Analytics Lab Friedrich-Alexander-Universität Erlangen-Nürnberg Germany STABILO International GmbH Germany

Most online handwriting recognition systems require the use of specific writing surfaces to extract positional data. In this paper we present a online handwriting recognition system for word recognition which is based on inertial measurement units (IMUs) for digitizing text written on paper. This is obtained by means of a sensor-equipped pen that provides acceleration, angular velocity, and magnetic forces streamed via Bluetooth. Our model combines convolutional and bidirectional LSTM networks, and is trained with the Connectionist Temporal Classification loss that allows the interpretation of raw sensor data into words without the need of sequence segmentation. We use a dataset of words collected using multiple sensor-enhanced pens and evaluate our model on distinct test sets of seen and unseen words achieving a character error rate of 17.97% and 17.08%, respectively, without the use of a dictionary or language model. © 2021, CC BY.

关键词： Statistical tests

Latent space conditioning on generative adversarial networks

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Lopez, Ricard Durall Ho, Kalun Pfreundt, Franz-Josef Keuper, Janis Fraunhofer ITWM Germany IWR University of Heidelberg Germany Fraunhofer Center Machine Learning Germany Data and Web Science Group University of Mannheim Germany Institute for Machine Learning and Analytics Offenburg University Germany

Generative adversarial networks are the state of the art approach towards learned synthetic image generation. Although early successes were mostly unsupervised, bit by bit, this trend has been superseded by approaches based on labelled data. These supervised methods allow a much finer-grained control of the output image, offering more flexibility and stability. Nevertheless, the main drawback of such models is the necessity of annotated data. In this work, we introduce an novel framework that benefits from two popular learning techniques, adversarial training and representation learning, and takes a step towards unsupervised conditional GANs. In particular, our approach exploits the structure of a latent space (learned by the representation learning) and employs it to condition the generative model. In this way, we break the traditional dependency between condition and label, substituting the latter by unsupervised features coming from the latent space. Finally, we show that this new technique is able to produce samples on demand keeping the quality of its supervised counterpart. © 2020, CC BY.

关键词： Generative adversarial networks

Unmasking DeepFakes with simple features

学校读者我要写书评

暂无评论

arXiv 2019年

作者： Durall, Ricard Keuper, Margret Pfreundt, Franz-Josef Keuper, Janis Fraunhofer ITWM Germany IWR University of Heidelberg Germany Fraunhofer Center Machine Learning Germany Data and Web Science Group University Mannheim Germany Institute for Machine Learning and Analytics Offenburg University Germany

Deep generative models have recently achieved impressive results for many real-world applications, successfully generating high-resolution and diverse samples from complex data sets. Due to this improvement, fake digital contents have proliferated growing concern and spreading distrust in image content, leading to an urgent need for automated ways to detect these AI-generated fake images. Despite the fact that many face editing algorithms seem to produce realistic human faces, upon closer examination, they do exhibit artifacts in certain domains which are often hidden to the naked eye. In this work, we present a simple way to detect such fake face images - so-called DeepFakes. Our method is based on a classical frequency domain analysis followed by a basic classifier. Compared to previous systems, which need to be fed with large amounts of labeled data, our approach showed very good results using only a few annotated training samples and even achieved good accuracies in fully unsupervised scenarios. For the evaluation on high resolution face images, we combined several public data sets of real and fake faces into a new benchmark: Faces-HQ. Given such high-resolution images, our approach reaches a perfect classification accuracy of 100% when it is trained on as little as 20 annotated samples. In a second experiment, in the evaluation of the medium-resolution images of the CelebA data set, our method achieves 100% accuracy supervised and 96% in an unsupervised setting. Finally, evaluating a low-resolution video sequences of the FaceForensics++ data set, our method achieves 91% accuracy detecting manipulated videos. Copyright © 2019, The Authors. All rights reserved.

关键词： Frequency domain analysis

machine learning Techniques for Stress Prediction in Working Employees

学校读者我要写书评

暂无评论

Machine Learning Techniques for Stress Prediction in Working...

IEEE International Conference on Computational Intelligence and Computing Research

作者： U Srinivasulu Reddy Aditya Vivek Thota A Dharun Department of Computer Applications Machine Learning and Data Analytics Lab National Institute of Technology Trichy ECE National Institute of Technology Trichy

ISBN: (纸本)9781538615096

Stress disorders are a common issue among working IT professionals in the industry today. With changing lifestyle and work cultures, there is an increase in the risk of stress among the employees. Though many industries and corporates provide mental health related schemes and try to ease the workplace atmosphere, the issue is far from control. In this paper, we would like to apply machine learning techniques to analyze stress patterns in working adults and to narrow down the factors that strongly determine the stress levels. Towards this, data from the OSMI mental health survey 2017 responses of working professionals within the tech-industry was considered. Various machine learning techniques were applied to train our model after due data cleaning and preprocessing. The accuracy of the above models was obtained and studied comparatively. Boosting had the highest accuracy among the models implemented. By using Decision Trees, prominent features that influence stress were identified as gender, family history and availability of health benefits in the workplace. With these results, industries can now narrow down their approach to reduce stress and create a much comfortable workplace for their employees.

关键词： data handling decision trees DP management employee welfare ergonomics gender issues health care learning (artificial intelligence) occupational stress organisational aspects personnel psychology risk analysis computer installation management Decision Trees Organizational aspects gender issue data processing stress at work Ergonomics Psychology Risk Management risk analysis Personnel Employees Health Care data cleaning learning methods