检索结果-内蒙古大学图书馆

A neural model for joint document and snippet ranking in question answering for large document collections

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Pappas, Dimitris Androutsopoulos, Ion Department of Informatics Athens University of Economics and Business Greece Institute for Language and Speech Processing Research Center 'Athena' Greece

Question answering (QA) systems for large document collections typically use pipelines that (i) retrieve possibly relevant documents, (ii) re-rank them, (iii) rank paragraphs or other snippets of the top-ranked documents, and (iv) select spans of the top-ranked snippets as exact answers. Pipelines are conceptually simple, but errors propagate from one component to the next, without later components being able to revise earlier decisions. We present an architecture for joint document and snippet ranking, the two middle stages, which leverages the intuition that relevant documents have good snippets and good snippets come from relevant documents. The architecture is general and can be used with any neural text relevance ranker. We experiment with two main instantiations of the architecture, based on POSIT-DRMM (PDRMM) and a BERT-based ranker. Experiments on biomedical data from BIOASQ show that our joint models vastly outperform the pipelines in snippet retrieval, the main goal for QA, with fewer trainable parameters, also remaining competitive in document retrieval. Furthermore, our joint PDRMM-based model is competitive with BERT-based models, despite using orders of magnitude fewer parameters. These claims are also supported by human evaluation on two test batches of BIOASQ. To test our key findings on another dataset, we modified the Natural Questions dataset so that it can also be used for document and snippet retrieval. Our joint PDRMM-based model again outperforms the corresponding pipeline in snippet retrieval on the modified Natural Questions dataset, even though it performs worse than the pipeline in document retrieval. We make our code and the modified Natural Questions dataset publicly *** Codes 68P20, 68P10, 68T50, 68T07 © 2021, CC BY.

关键词： Pipelines

Sample-Efficient Unsupervised Domain Adaptation of speech Recognition Systems A case study for Modern Greek

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Paraskevopoulos, Georgios Kouzelis, Theodoros Rouvalis, Georgios Katsamanis, Athanasios Katsouros, Vassilis Potamianos, Alexandros The Graduate School of ECE National Technical University of Athens Athens Greece The Institute for Speech and Language Processing Athena Research Center Athens Greece The Faculty of ECE National Technical University of Athens Athens Greece

Modern speech recognition systems exhibits rapid performance degradation under domain shift. This issue is especially prevalent in data-scarce settings, such as low-resource languages, where diversity of training data is limited. In this work we propose M2DS2, a simple and sample-efficient finetuning strategy for large pretrained speech models, based on mixed source and target domain self-supervision. We find that including source domain self-supervision stabilizes training and avoids mode collapse of the latent representations. For evaluation, we collect HParl, a 120 hour speech corpus for Greek, consisting of plenary sessions in the Greek Parliament. We merge HParl with two popular Greek corpora to create GREC-MD, a testbed for multi-domain evaluation of Greek ASR systems. In our experiments we find that, while other Unsupervised Domain Adaptation baselines fail in this resource-constrained environment, M2DS2 yields significant improvements for cross-domain adaptation, even when a only a few hours of in-domain audio are available. When we relax the problem in a weakly supervised setting, we find that independent adaptation for audio using M2DS2 and language using simple LM augmentation techniques is particularly effective, yielding word error rates comparable to the fully supervised baselines. © 2022, CC BY.

关键词： speech recognition

Sample-Efficient Unsupervised Domain Adaptation of speech Recognition Systems: A case study for Modern Greek

学校读者我要写书评

暂无评论

TechRxiv

TechRxiv 2022年

关键词： speech recognition

Neural Sign Reenactor: Deep Photorealistic Sign language Retargeting

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Tze, Christina O. Filntisis, Panagiotis P. Dimou, Athanasia-Lida Roussos, Anastasios Maragos, Petros School of Electrical & Computer Engineering National Technical University of Athens Greece Greece College of Engineering Mathematics and Physical Sciences University of Exeter United Kingdom Institute for Language and Speech Processing Athena R.C. Greece Greece

In this paper, we introduce a neural rendering pipeline for transferring the facial expressions, head pose, and body movements of one person in a source video to another in a target video. We apply our method to the challenging case of Sign language videos: given a source video of a sign language user, we can faithfully transfer the performed manual (e.g. handshape, palm orientation, movement, location) and non-manual (e.g. eye gaze, facial expressions, mouth patterns, head, and body movements) signs to a target video in a photo-realistic manner. Our method can be used for Sign language Anonymization, Sign language Production (synthesis module), as well as for reenacting other types of full body activities (dancing, acting performance, exercising, etc.). We conduct detailed qualitative and quantitative evaluations and comparisons, which demonstrate the particularly promising and realistic results that we obtain and the advantages of our method over existing approaches. © 2022, CC BY.

关键词： Eye movements

The myth of signing avatars 1

学校读者我要写书评

暂无评论

The myth of signing avatars

1st International Workshop on Automatic Translation for Signed and Spoken languages, AT4SSL 2021

作者： Wolfe, Rosalee McDonald, John C. Efthimiou, Eleni Fontinea, Evita Picron, Frankie Van Landuyt, Davy Sioen, Tina Braffort, Annelies Filhol, Michael Ebling, Sarah Hanke, Thomas Krausneker, Verena Institute for Language and Speech Processing Athena RC Greece School of Computing DePaul University Chicago United States European Union of the Deaf Brussels Belgium Laboratoire Interdisciplinaire des Sciences du Numérique Orsay France Department of Computational Linguistics University of Zurich Switzerland Institut für Deutsche Gebärdensprache Universität Hamburg Germany Institut für Sprachwissenschaft Universität Wien Vienna Austria

Development of automatic translation between signed and spoken languages has lagged behind the development of automatic translation between spoken languages, but it is a common misperception that extending machine translation techniques to include signed languages should be a straightforward process. A contributing factor is the lack of an acceptable method for displaying sign language apart from interpreters on video. This position paper examines the challenges of displaying a signed language as a target in automatic translation, analyses the underlying causes and suggests strategies to develop display technologies that are acceptable to sign language communities. © 2021 Proceedings of the 1st International Workshop on Automatic Translation for Signed and Spoken languages, AT4SSL 2021. All rights reserved.

关键词： Translation (languages)

Visual speech-Aware Perceptual 3D Facial Expression Reconstruction from Videos

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Filntisis, Panagiotis P. Retsinas, George Paraperas-Papantoniou, Foivos Katsamanis, Athanasios Roussos, Anastasios Maragos, Petros School of Electrical & Computer Engineering National Technical University of Athens Greece Greece College of Engineering Mathematics and Physical Sciences University of Exeter United Kingdom Imperial College London United Kingdom Institute for Language and Speech Processing Athena R.C. Greece

The recent state of the art on monocular 3D face reconstruction from image data has made some impressive advancements, thanks to the advent of Deep Learning. However, it has mostly focused on input coming from a single RGB image, overlooking the following important factors: a) Nowadays, the vast majority of facial image data of interest do not originate from single images but rather from videos, which contain rich dynamic information. b) Furthermore, these videos typically capture individuals in some form of verbal communication (public talks, teleconferences, audiovisual human-computer interactions, interviews, monologues/dialogues in movies, etc). When existing 3D face reconstruction methods are applied in such videos, the artifacts in the reconstruction of the shape and motion of the mouth area are often severe, since they do not match well with the speech audio. To overcome the aforementioned limitations, we present the first method for visual speech-aware perceptual reconstruction of 3D mouth expressions. We do this by proposing a "lipread" loss, which guides the fitting process so that the elicited perception from the 3D reconstructed talking head resembles that of the original video footage. We demonstrate that, interestingly, the lipread loss is better suited for 3D reconstruction of mouth movements compared to traditional landmark losses, and even direct 3D supervision. Furthermore, the devised method does not rely on any text transcriptions or corresponding audio, rendering it ideal for training in unlabeled datasets. We verify the efficiency of our method through exhaustive objective evaluations on three large-scale datasets, as well as subjective evaluation with two web-based user studies. © 2022, CC BY.

关键词： Human computer interaction

FedPOIRec: Privacy preserving federated POI recommendation with social influence

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Perifanis, Vasileios Drosatos, George Stamatelatos, Giorgos Efraimidis, Pavlos S. Institute for Language and Speech Processing Athena Research Center Xanthi Kimmeria67100 Greece Department of Electrical and Computer Engineering Democritus University of Thrace Xanthi Kimmeria67100 Greece

With the growing number of Location-Based Social Networks, privacy preserving location prediction has become a primary task for helping users discover new points-of-interest (POIs). Traditional systems consider a centralized approach that requires the transmission and collection of users' private data. In this work, we present FedPOIRec, a privacy preserving federated learning approach enhanced with features from users' social circles for top-N POI recommendations. First, the FedPOIRec framework is built on the principle that local data never leave the owner's device, while the local updates are blindly aggregated by a parameter server. Second, the local recommenders get personalized by allowing users to exchange their learned parameters, enabling knowledge transfer among friends. To this end, we propose a privacy preserving protocol for integrating the preferences of a user's friends after the federated computation, by exploiting the properties of the CKKS fully homomorphic encryption scheme. To evaluate FedPOIRec, we apply our approach into five real-world datasets using two recommendation models. Extensive experiments demonstrate that FedPOIRec achieves comparable recommendation quality to centralized approaches, while the social integration protocol incurs low computation and communication overhead on the user side. Copyright © 2021, The Authors. All rights reserved.

关键词： Knowledge management

Exploiting Extended Reality Technologies for Educational Microscopy 17th

学校读者我要写书评

暂无评论

Exploiting Extended Reality Technologies for Educational Mic...

17th International Conference on Virtual Reality and Augmented Reality, EuroVR 2020

作者： Theodoropoulou, Helena G. Kiourt, Chairi Lalos, Aris S. Koutsoudis, Anestis Paxinou, Evgenia Kalles, Dimitris Pavlidis, George Industrial Systems Institute Athena-Research and Innovation Center in Information Communication and Knowledge Technologies Patra Greece Institute for Language and Speech Processing Athena-Research and Innovation Center in Information Communication and Knowledge Technologies Xanthi Greece School of Science and Technology Hellenic Open University Patra Greece

ISBN: (纸本)9783030626549

Exploiting extended reality technologies in laboratory training enhances both teaching and learning experiences. It complements the existing traditional learning/teaching methods related to science, technology, engineering, arts and mathematics. In this work, we use extended reality technologies to create an interactive learning environment with dynamic educational content. The proposed learning environment can be used by students of all levels of education, to facilitate laboratory-based understanding of scientific concepts. We introduce a low-cost and user-friendly multi-platform system for mobile devices which, when coupled with edutainment dynamics, simulation, extended reality and natural hand movements sensing technologies such as hand gestures with virtual triggers, is expected to engage users and prepare them efficiently for the actual on-site laboratory experiments. The proposed system is evaluated by a group of experts and the results are analyzed in detail, indicating the positive attitude of the evaluators towards the adoption of the proposed system in laboratory educational procedures. We conclude the paper by highlighting the capabilities of extended reality and dynamic content management in educational microscopy procedures. © 2020, Springer Nature Switzerland AG.

关键词： Laboratories

Air-Writing Recognition using Deep Convolutional and Recurrent Neural Network Architectures

学校读者我要写书评

暂无评论

Air-Writing Recognition using Deep Convolutional and Recurre...

International Workshop on Frontiers in Handwriting Recognition

作者： Grigoris Bastas Kosmas Kritsis Vassilis Katsouros Institute for Language and Speech Processing (ILSP) Athena R.C. Athens Greece

ISBN: (数字)9781728199665

ISBN: (纸本)9781728199672

In this paper, we explore deep learning architectures applied to the air-writing recognition problem where a person writes text freely in the three dimensional space. We focus on handwritten digits, namely from 0 to 9, which are structured as multidimensional time-series acquired from a Leap Motion Controller (LMC) sensor. We examine both dynamic and static approaches to model the motion trajectory. We train and compare several state-of-the-art convolutional and recurrent architectures. Specifically, we employed a Long Short-Term Memory (LSTM) network and also its bidirectional counterpart (BLSTM) in order to map the input sequence to a vector of fixed dimensionality, which is subsequently passed to a dense layer for classification among the targeted air-handwritten classes. In the second architecture we adopt 1D Convolutional Neural Networks (CNNs) to encode the input features before feeding them to an LSTM neural network (CNN-LSTM). The third architecture is a Temporal Convolutional Network (TCN) that uses dilated causal convolutions. Finally, a deep CNN architecture for automating the feature learning and classification from raw input data is presented. The performance evaluation has been carried out on a dataset of 10 participants, who wrote each digit at least 10 times, resulting in almost 1200 examples.

关键词： Trajectory Writing Computer architecture Character recognition Hidden Markov models Two dimensional displays Tracking