The effectiveness of GANs in producing images according to a specific visual domain has shown potential in unsupervised domain adaptation. Source labeled images have been modified to mimic target samples for training classifiers in the target domain, and inverse mappings from the target to the source domain have also been evaluated, without new image generation.In this paper we aim at getting the best of both worlds by introducing a symmetric mapping among domains. We jointly optimize bi-directional image transformations combining them with target self-labeling. We define a new class consistency loss that aligns the generators in the two directions, imposing to preserve the class identity of an image passing through both domain mappings. A detailed analysis of the reconstructed images, a thorough ablation study and extensive experiments on six different settings confirm the power of our approach.

ABSTRACT:The combined use of high-resolution digital images taken from ground as well as from RPAS (Remotely Piloted Aircraft Systems) have significantly increased the potential of close range digital photogrammetry applications in Cultural Heritage surveying and modeling. It is in fact possible, thanks to SfM (Structure from Motion), to simultaneously process great numbers of aerial and terrestrial images for the production of a dense point cloud of an object. In order to analyze the accuracy of results, we started numerous tests based on the comparison between 3D digital models of a monumental complex realized by the integration of aerial and terrestrial photogrammetry and an accurate TLS (Terrestrial Laser Scanner) reference model of the same object. A lot of digital images of a renaissance castle, assumed as test site, have been taken both by ground level and by RPAS at different distances and flight altitudes and with different flight patterns. As first step of the experimentation, the images were previously processed with Agisoft PhotoScan, one of the most popular photogrammetric software. The comparison between the photogrammetric DSM of the monument and a TLS reference one was carried out by evaluating the average deviation between the points belonging to the two entities, both globally and locally, on individual façades and architectural elements (sections and particular). In this paper the results of the first test are presented. A good agreement between photogrammetric and TLS digital models of the castle is pointed out.

The no-visitor policies endorsed by healthcare organizations to limit COVID-19 virus risk exposure have unfortunately contributed to the isolation of patients further exacerbating distress in relatives and frontline healthcare workers. To contrast such effects, many healthcare institutions have adopted technology-based solutions helping patients and families communicate online through the aid of virtual devices. To date, no study has investigated whether facilitating patient-family videocalls would mitigate distress levels in frontline healthcare professionals. Caring for emotional needs of patients by re-establishing affiliative connections interrupted by the pandemic through patient-family videocalls is expected to mitigate distress in engaged healthcare workers as an example of a tend-and-befriend response to stress caused by the pandemic. We tested this hypothesis in a cross-sectional study conducted during 1-30 June 2020, involving 209 healthcare workers (nurses = 146; physicians = 63) engaged in the COVID-19 frontline in Italy. Half of participants in our sample (n = 107) had assisted efforts aimed at connecting patients remotely with families through videocalls. Psychological distress measures included symptoms of burnout, post-traumatic stress, anxiety, depression, and difficulty in sleep and wakefulness. Partially in line with our expectations we found a modulation effect specific for professional category: nurses assisting patient-family videocalls reported significantly lower levels of distress and a better quality of wakefulness compared to those who did not, whereas physicians reported higher levels of distress during such virtual communications. We interpret these findings from the perspective of patient-family communication and differences in skills and training between nurses and physicians. These findings highlight that technology-based solutions aimed at reducing barriers and alleviating distress in healthcare settings should be promoted in concert with skill enhancement training for healthcare professionals especially in terms of communicating online and communicating difficult topics with patients and families.

ABSTRACT:In order to analyze the potential as well as the limitations of low-cost RPAS photogrammetric systems for architectural cultural heritage reconstruction, some tests were performed by a small RPAS equipped with an ultralight camera. The tests were carried out in a site of remarkable historical interest. A great amount of images were taken with camera's optical axis in vertical and oblique position. Images were processed by the commercial software PhotoScan of Agisoft and numerous models were realized, each of them was compared with an accurate TLS model used as a reference. The test, despite some problems found, has provided good results in terms of accuracy (average error <2cm) and reliability.

The sudden algal bloom in shallow water may be a serious problem for sea coastal economy based on clams farming because it leads quickly to anoxia conditions with the consequent death of the molluscs. In order to detect the rise of algae, normally satellite remote sensing is used, exploiting the higher response in the near infrared wavelengths. A recent progress in monitoring this phenomenon derives from the availability of unmanned aerial vehicles (UAVs) equipped with lightweight multispectral cameras. Such technique makes it possible to acquire detailed spectral information with narrow bands attaining an assessment of the algal bloom at both high geometric and radiometric resolutions. In this work, we tested the MicaSense RedEdge-M multispectral camera mounted on a DJI Phantom 3 Professional aircraft to map submerged seaweeds and assess their evolution with particular regard to the importance of the radiometric calibration of raw imageries using a Downwelling Light Sensor (DLS) and a known reflectance panel. The case study is the lagoon of Goro (Northern Adriatic Sea, Italy), a crucial environment for the clams farming in the Emilia-Romagna region. Digital images acquired in two subsequent flights were processed with either Agisoft PhotoScan Professional and Pix4D Mapper Pro varying the calibration strategies. After a pre-analysis, we applied two different approaches for the seaweed detection: NDVI and maximum likelihood classification. All the tests performed in this study confirm that the monitoring over time with a multispectral lightweight camera mounted on a UAV is possible, but also that by applying proper radiometric corrections, most accurate and reliable results can be achieved.