Found 9883 publications. Showing page 178 of 396:
2002
2023
Identifying knowledge gaps in physical preservation of cultural heritage prior to 2020. NILU OR
NILU - the Norwegian Institute for Air Research and NIKU - the Norwegian Institute for Cultural Heritage Research has upon application received funding from the Directorate for Cultural Heritage in Norway to prepare a memorandum that identifies knowledge gaps in Norway related to physical preservation of cultural heritage.
2013
2015
2011
Identifying trends in the ocean wave climate by time series analyses of significant wave heightdata.
2013
2020
2020
2005
Cross-modal representation learning aims to learn a shared representation space where data from multiple modalities can be effectively compared, fused, and understood. This paper investigates the role of increased diversity in the similarity score matrix in enhancing the performance of the CLIP (Contrastive Language-Image Pretraining), a multi-modal learning model that establishes a connection between images and text within a joint embedding space. Two transforming approaches, sine and sigmoid (including two versions), are incorporated into the CLIP model to amplify larger values and diminish smaller values within the similarity matrix (logits). Hardware limitations are addressed using a more compact text encoder (DistilBERT) and a pre-trained ResNet50 image encoder. The proposed adaptations are evaluated on various benchmarks, including image classification and image/text retrieval tasks, using 10 benchmark datasets such as Food101, Flickr30k, and COCO. The performance of the adapted models is compared to the base CLIP model using Accuracy, mean per class, and Recall@k metrics. The results demonstrate improvements in Accuracy (up to 5.32% enhancement for the PatchCamelyon dataset), mean per class (up to 14.48% enhancement for the FGVCAircraft dataset), and retrieval precision (with an increase of up to 45.20% in Recall@1 for the COCO dataset), compared to the baseline algorithm (CLIP).
IEEE (Institute of Electrical and Electronics Engineers)
2023
2012
2008
2015
2017
Impact assessment for emissions to air from a planned aluminium smelter in Reydarfjordur, Iceland. NILU OR
2001
2003
2004
2012
Impact of 2020 COVID-19 lockdowns on particulate air pollution across Europe
To fight against the first wave of coronavirus disease 2019 (COVID-19) in 2020, lockdown measures were implemented in most European countries. These lockdowns had well-documented effects on human mobility. We assessed the impact of the lockdown implementation and relaxation on air pollution by comparing daily particulate matter (PM), nitrogen dioxide (NO2) and ozone (O3) concentrations, as well as particle number size distributions (PNSDs) and particle light absorption coefficient in situ measurement data, with values that would have been expected if no COVID-19 epidemic had occurred at 28 sites across Europe for the period 17 February–31 May 2020. Expected PM, NO2 and O3 concentrations were calculated from the 2020 Copernicus Atmosphere Monitoring Service (CAMS) ensemble forecasts, combined with 2019 CAMS ensemble forecasts and measurement data. On average, lockdown implementations did not lead to a decrease in PM2.5 mass concentrations at urban sites, while relaxations resulted in a +26 ± 21 % rebound. The impacts of lockdown implementation and relaxation on NO2 concentrations were more consistent (−29 ± 17 and +31 ± 30 %, respectively). The implementation of the lockdown measures also induced statistically significant increases in O3 concentrations at half of all sites (+13 % on average). An enhanced oxidising capacity of the atmosphere could have boosted the production of secondary aerosol at those places. By comparison with 2017–2019 measurement data, a significant change in the relative contributions of wood and fossil fuel burning to the concentration of black carbon during the lockdown was detected at 7 out of 14 sites. The contribution of particles smaller than 70 nm to the total number of particles significantly also changed at most of the urban sites, with a mean decrease of −7 ± 5 % coinciding with the lockdown implementation. Our study shows that the response of PM2.5 and PM10 mass concentrations to lockdown measures was not systematic at various sites across Europe for multiple reasons, the relationship between road traffic intensity and particulate air pollution being more complex than expected.
2023