Abstract
The Feedfirst Intelligent Monitoring System is a novel tool for intelligent monitoring of fish nurseries in aquaculture scenarios, mainly focusing on monitoring three essential items: water quality control, biomass estimation, and automated feeding. The system is based on machine vision techniques for fish larvae population size detection, and larvae biomass estimation is monitored through size measurement. We also show that the perception-actuation loop in automated fish tanks can be closed by using the vision system output to influence feeding procedures. The proposed solution was tested in a real tank in an aquaculture setting with real-time performance and logging capabilities.

Abstract
Precancerous conditions such as intestinal metaplasia (IM) have a key role in gastric cancer development and can be detected during endoscopy. During upper gastrointestinal endoscopy (UGIE), misdiagnosis can occur due to technical and human factors or by the nature of the lesions, leading to a wrong diagnosis which can result in no surveillance/treatment and impairing the prevention of gastric cancer. Deep learning systems show great potential in detecting precancerous gastric conditions and lesions by using endoscopic images and thus improving and aiding physicians in this task, resulting in higher detection rates and fewer operation errors. This study aims to develop deep learning algorithms capable of detecting IM in UGIE images with a focus on model explainability and interpretability. In this work, white light and narrow-band imaging UGIE images collected in the Portuguese Institute of Oncology of Porto were used to train deep learning models for IM classification. Standard models such as ResNet50, VGG16 and InceptionV3 were compared to more recent algorithms that rely on attention mechanisms, namely the Vision Transformer (ViT), trained in 818 UGIE images (409 normal and 409 IM). All the models were trained using a 5-fold cross-validation technique and for validation, an external dataset will be tested with 100 UGIE images (50 normal and 50 IM). In the end, explainability methods (Grad-CAM and attention rollout) were used for more clear and more interpretable results. The model which performed better was ResNet50 with a sensitivity of 0.75 (±0.05), an accuracy of 0.79 (±0.01), and a specificity of 0.82 (±0.04). This model obtained an AUC of 0.83 (±0.01), where the standard deviation was 0.01, which means that all iterations of the 5-fold cross-validation have a more significant agreement in classifying the samples than the other models. The ViT model showed promising performance, reaching similar results compared to the remaining models. © 2023, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

Abstract

Abstract
As the availability of computational power and communication technologies increases, Humans and systems are able to tackle increasingly challenging decision problems. Taking decisions over incomplete visions of a situation is particularly challenging and calls for a set of intertwined skills that must be put into place under a clear rationale. This work addresses how to deliver autonomous decisions for the management of a public street lighting network, to optimize energy consumption without compromising light quality patterns. Our approach is grounded in an holistic methodology, combining semantic and Artificial Intelligence principles to define methods and artefacts for supporting decisions to be taken in the context of an incomplete domain. That is, a domain with absence of data and of explicit domain assertions.

Abstract
The sense of presence is an important aspect of experiences in Virtual Reality (VR), an emerging technology in education, leading this construct to be increasingly researched in parallel to learning purposes. However, there is not a consensus in the literature on the outcomes of this association. Aiming to outline a panorama in this regard, a systematic literature review was conducted, with a comprehensive analysis of 140 primary studies recovered from five worldwide databases. The analysis shows an overview of 24 years of areas, factors, and methodological approaches that seem to be more inclined to benefit from the sense of presence toward learning purposes. We contribute to the advancement of state of the art by providing an understanding of the relationship among these variables, identifying potential ways to benefit from the sense of presence to further leverage the use of VR for learning purposes.

Abstract