Results 11 to 20 of about 544,807 (273)
FinEst BERT and CroSloEngual BERT [PDF]
, 2020 Large pretrained masked language models have become state-of-the-art solutions for many NLP problems. The research has been mostly focused on English language, though. While massively multilingual models exist, studies have shown that monolingual models produce much better results.
Ulčar, Matej, Robnik-Šikonja, Marko
openaire +3 more sources
Journal of Road Engineering, 2023 Microbial-induced calcium carbonate precipitation is a promising technology for self-healing concrete due to its capability to seal microcracks. The main goal of this study was to evaluate the effects of adding hydrogel-encapsulated bacteria on the ...
Ricardo Hungria +2 more
doaj +1 more source
Seasonal variation of aerosol water uptake and its impact on the direct radiative effect at Ny-Ålesund, Svalbard [PDF]
Atmospheric Chemistry and Physics, 2014 In this study we investigated the impact of water uptake by aerosol particles in ambient atmosphere on their optical properties and their direct radiative effect (ADRE, W m−2) in the Arctic at Ny-Ålesund, Svalbard, during 2008.
N. Rastak +10 more
doaj +1 more source
A Lite Romanian BERT: ALR-BERT
Computers, 2022 Large-scale pre-trained language representation and its promising performance in various downstream applications have become an area of interest in the field of natural language processing (NLP). There has been huge interest in further increasing the model’s size in order to outperform the best previously obtained performances.
Dragoş Constantin Nicolae +2 more
openaire +3 more sources
Importance of aerosol composition and mixing state for cloud droplet activation over the Arctic pack ice in summer [PDF]
Atmospheric Chemistry and Physics, 2015 Concentrations of cloud condensation nuclei (CCN) were measured throughout an expedition by icebreaker around the central Arctic Ocean, including a 3 week ice drift operation at 87° N, from 3 August to 9 September 2008.
C. Leck, E. Svensson
doaj +1 more source
Multi-mode deep auto-encoder recommendation model for fusion of text information
Xi'an Gongcheng Daxue xuebao, 2021 A new recommendation model is proposed to solve the problems of data sparsity and deep semantic information learning when using score information as auxiliary recommendation.
Jinguang CHEN, Xinyi XU, Ganglong FAN
doaj +1 more source
Micron-BERT: BERT-Based Facial Micro-Expression Recognition
2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023 Micro-expression recognition is one of the most challenging topics in affective computing. It aims to recognize tiny facial movements difficult for humans to perceive in a brief period, i.e., 0.25 to 0.5 seconds. Recent advances in pre-training deep Bidirectional Transformers (BERT) have significantly improved self-supervised learning tasks in computer
Nguyen, Xuan-Bac +5 more
openaire +2 more sources
Catchment-scale dissolved carbon concentrations and export estimates across six subarctic streams in northern Sweden [PDF]
Biogeosciences, 2014 Climatic change is currently enhancing permafrost thawing and the flow of water through the landscape in subarctic and arctic catchments, with major consequences for the carbon export to aquatic ecosystems.
R. Giesler +7 more
doaj +1 more source
AUBER: Automated BERT regularization
PLOS ONE, 2021 How can we effectively regularize BERT? Although BERT proves its effectiveness in various NLP tasks, it often overfits when there are only a small number of training instances. A promising direction to regularize BERT is based on pruning its attention heads with a proxy score for head importance. However, these methods are usually suboptimal since they
Hyun Dong Lee, Seongmin Lee, U. Kang
openaire +5 more sources
BERT-CoQAC: BERT-Based Conversational Question Answering in Context [PDF]
, 2021 As one promising way to inquire about any particular information through a dialog with the bot, question answering dialog systems have gained increasing research interests recently. Designing interactive QA systems has always been a challenging task in natural language processing and used as a benchmark to evaluate a machine's ability of natural ...
Zaib, Munazza +5 more
openaire +2 more sources