In recent years, the EU has observed a decrease in the stocks of certain fish species due to unrestricted fishing. To combat the problem, many fisheries are investigating how to automatically estimate the catch size and composition using sensors onboard the vessels. Yet, measuring the size of fish in marine imagery is a difficult task. The images generally suffer from complex conditions caused by cluttered fish, motion blur and dirty sensors. In this thesis, we propose a novel method for automatic measurement of fish size that can enable measuring both visible and occluded fish. We use a Mask R-CNN to segment the visible regions of the fish, and then fill in the shape of the occluded fish using a U-Net. We train the U-Net to perform shape completion in a semi-supervised manner, by simulating occlusions on an open-source fish dataset. Different to previous shape completion work, we teach the U-Net when to fill in the shape and not by including a small portion of fully visible fish in the input training data. Our results show that our proposed method succeeds to fill in the shape of the synthetically occluded fish as well as of some of the cluttered fish in real marine imagery. We achieve an mIoU score of 93.9 % on 1 000 synthetic test images and present qualitative results on real images captured onboard a fishing vessel. The qualitative results show that the U-Net can fill in the shapes of lightly occluded fish, but struggles when the tail fin is hidden and only parts of the fish body is visible. This task is difficult even for a human, and the performance could perhaps be increased by including the fish appearance in the shape completion task. The simulation-to-reality gap could perhaps also be reduced by finetuning the U-Net on some real occlusions, which could increase the performance on the heavy occlusions in the real marine imagery.

No description available.

Equipment and machines can be exposed to very high temperatures in the steel mill industry. One particularly critical part is the ladles used to hold and pour molten iron into mouldings. A refractory lining is used as an insulation layer between the outer steel shell and the molten iron to protect the ladle from the hot iron. Over time, or if the lining is not completely cured, the lining wears out or can potentially fail. Such a scenario can lead to a breakout of molten iron, which can cause damage to equipment and, in the worst case, workers. Previous work analyses how critical areas can be identified in a proactive matter. Using thermal imaging, the failing spots on the lining could show as high-temperature areas on the outside steel shell. The idea is that the outside temperature corresponds to the thickness of the insulating lining. The detection of these spots is identified when temperatures over a given threshold are registered within the thermal camera's field of view. The images must then be manually analyzed over time, to follow the progression of a detected spot. The existing solution is also prone to the background noise of other hot objects.  This thesis proposes an initial step to automate monitoring the health of refractory lining in steel ladles. The report will investigate the usage of Instance Segmentation to isolate the ladle from its background. Thus, reducing false alarms and background noise in an autonomous monitoring setup. The model training is based on Mask R-CNN on our own thermal images, with pre-trained weights from visual images. Detection is done on two classes: open or closed ladle. The model proved reasonably successful on a small dataset of 1000 thermal images. Different models were trained with and without augmentation, pre-trained weights as well multi-phase fine-tuning. The highest mAP of 87.5\% was achieved on a pre-trained model with image augmentation without fine-tuning. Though it was not tested in production, temperature readings could lastly be extracted on the segmented ladle, decreasing the risk of false alarms from background noise.

Introduktion:Pandemier hotar folkhälsan och kan ge allvarliga konsekvenser på individ och befolkningsnivå. Munskydd ses som en förebyggande åtgärd för att skydda sin egen hälsa och förhindrar smittspridning till medmänniskor i samhället. Patienter i riskgrupp var goda samhällsmedborgare och visade en god efterlevnad till munskyddsanvändning. Syfte: Att undersöka riskgruppspatienters kunskap och upplevelser av användning och hälsoinformation om munskydd under Covid 19-pandemin i Sverige. Metod:En kvalitativ metod användes och datainsamlingen genomfördes genom semistrukturerade telefonintervjuer. Deltagarna (n=7) var riskgruppspatienter av svenskt ursprung och använde munskydd under covidpandemin. Kvalitativ innehållsanalys användes enligt Graneheim & Lundman (2004) för att analysera data. Temaidentifieringen genomfördes enligt Nutbeams(2000) beskrivning av tre dimensioner av health literacy. Resultat:Studiedeltagarna har varierande erfarenheter kring munskyddsanvändning. Under covidpandemin använde inte många i allmänheten munskydd, vilket resulterade i att riskgrupper upplevde att de skiljdes ut när de bar munskydd. De upplevde rädslor när de befann sig ibland folk som inte bar munskydd, vilket påverkade patienterna negativt. Bärandet av munskyddet försvårade kommunikationen med omgivningen. Riskgrupper som prioriterades av myndigheter var de riskgrupper som löpte störst risk att drabbas av en allvarlig sjukdomsutveckling av covid-19. Enligt studiedeltagarnas erfarenhet fanns inte information för resterande riskgrupper, vilket resulterade i att patienterna inte trodde att de tillhörde någon riskgrupp. Riskgruppspatienter har fått en ökad förståelse om virusets smittvägar och använder munskydd på ett korrekt sätt. Slutsats: Studiedeltagarna har en varierande erfarenhet kring munskydd men de önskade strängarerekommendationer kring munskyddsanvändning, baserat på deras upplevelser kringmunskydd samt deras goda kunskaper om munskyddets nytta på individ och samhällsnivå. / Introduction:Pandemics threaten public health and can have serious consequences at the individual and population level. Face masks are seen as a preventive measure to protect one's own health and prevent the spread of infection to other people in society. Patients in the risk group were good citizens and showed good compliance with mouth protection use and social distancing. Aim:To investigate risk group patients' knowledge and experiences of use and health information about mouth protection during the Covid 19 pandemic in Sweden. Methods:A qualitative method was used, and the data collection was carried out through semi-structured telephone interviews. The participants (n=7) were risk group patients of Swedish origin and used mouth protection during the covid pandemic. Qualitative content analysis was used according to Graneheim & Lundman (2004) to analyze the data. Theme identification was carried out according to Nutbeam's (2000) description of three dimensions of health literacy. Results:The study participants have varying experiences regarding face masks use. During the covid pandemic, many members of the public did not use masks, resulting in at-risk groups feeling singled out when wearing masks. This gave rise to negative feelings. They also experienced fears when they were around people who did not wear masks. Wearing the mask made communication with the environment difficult. At-risk groups that were prioritized by authorities were the groups that ran the greatest risk of being affected by a serious disease development from covid-19. According to the study participants belief there was no information for the remaining at-risk groups available, which resulted in the patients not believing that they belonged to any risk group. Risk group patients have gained an increased understanding of the virus's transmission routes and use mouth protection correctly. Conclusion:This study strengthens the perception of the study participants' mixed physical and psychological experiences related to face masks use. The study participants' desire for stricter recommendations regarding face masks use is due to their experiences with face masks and their good knowledge of the benefits of face masks on an individual and societal level.

Диссертационное исследование обращено к проблеме изучения языковой личности. В работе реализована концепция описания языковой личности как лингвокультурного типажа. Объект исследования – языковая личность шоумена Сергея Светлакова. На первом этапе исследования на основе текстового материала, выбранного из Национального корпуса русского языка, проведен анализ лингвокультурных типажей простака и шутника, лежащих в основе речевых масок С.Светлакова, выявлены понятийные, образные и оценочные характеристики типажей. На втором этапе исследования описаны речевые маски простака и шутника С. Светлакова в коммуникативно-прагматическом аспекте. В работе были проанализированы коммуникативные стратегии и тактики речевой маски весельчака, которую «надевает» С. Светлаков в программе «Прожекторперисхилтон», а также типичного просторечника в скетч-шоу «Наша Russia». / This dissertation considers the problem of studying a linguistic personality. The work describes the concept of linguistic personality as a linguistic-cultural type. The object of research is the linguistic personality of a showman Sergei Svetlakov. The first phase of the work bases on « Russian National corpus» texts. We analyze linguistic-cultural types of the simpleton and the joker underlying Sergei Svetlakov’s speech masks. We also identified conceptual, imaginative and evaluative characteristics of the types. The second chapter describes speech masks of Sergei Svetlakov in the communicative-pragmatic aspect. The work analyzes the communication strategies and tactics of speech mask humorist, which «wears» Sergei Svetlakov in the TV program «Prozhektorperishilton», and typical of the representative of the vernacular type of culture in the sketch show «Our Russia».

Vid idrott utomhus i minusgrader utsätts andningssystemet för stora påfrestningar, luften är kall och torr vilket skadar och irriterar luftvägarna. För att undvika det kan man använda fukt- och värmeväxlare (HME). Dagens produkter har dålig passform vilket leder till läckage av luft. Syftet med projektet är att fler vinteratleter ska vilja använda HME och därmed förebygga förekomsten av andningssjukdomar. De tre delmålen för projektet är att tillverka minst en prototyp, den tillverkade prototypen ska ha bättre passform än dagens produkter och minst en lösning för slemhantering ska presenteras. Med hjälp av litteraturstudier lades en kunskapsgrund inom området. Verktyg som branschanalys, intressentanalys och målspecifikation gjorde konceptgenereringen rikare. Största delen av projektet lades i prototyptillverkning. Två prototyper kunde till slut presenteras, Winterflow och Winterflow Pro. Det är två additivt tillverkade masker med tillhörande filter som konstruerats i ett datormodelleringsprogram. Prototyperna validerades med hjälp av användartester. Prototyperna upplevs för en del testpersoner ha bättre passform än dagens produkter, men ingen statistisk signifikant skillnad gick att bevisa. Målen och syftet uppfylldes delvis och en bredare grund för vidare forskning fastställdes. / When performing sports outdoors during sub-zero conditions the respiratory system is exposed to great stress, the air is cold and dry, which damages and irritates the airways. To avoid this, the athletes can use heat and moisture exchanging devices (HME). Today’s products have a poor fit, which leads to leakage of the air. The aim of the project is to increase the usage of HME for winter athletes and thereby prevent the occurrence of respiratory diseases. The three objectives for the project are to manufacture at least one prototype, the prototype shall have a better fit than today’s products and at least one solution for handling saliva matter shall be presented. By carrying out literature studies, a foundation of knowledge was laid in the area. With the help of different analysis and tools, the concept generation were richer, resulting in more solutions. The main part of the project was put into prototyping. Two prototypes could be presented at the end, Winterflow and Winterflow Pro. Two additive manufactured masks with associated filters constructed in a computer modelling program. The prototypes were validated by user tests. The prototypes were perceived by some testers to have a better fit than today’s products, but no statistically significance could be proven. The aim and objectives were partly met, and a broader foundation of knowledge for further research was established. / <p>Betygsdatum 2023-06-07</p>

Multi-modality fusion is an area of research that has shown promising results in the domain of 2D and 3D object detection. However, multi-modality fusion methods have largely not been utilized in the domain of instance segmentation. This master’s thesis investigated if multi-modality fusion methods can be applied to deep learning instance segmentation models to improve their performance on multi-modality data. The two multi-modality fusion methods presented, input extension and feature fusions, were applied to a two-stage instance segmentation model, Mask R-CNN, and a single-stage instance segmentation model, RTMDet. Models were trained on different variations of preprocessed RGBD and ToF data provided by SICK IVP, as well as RGBD data from the publicly available NYUDepth dataset. The master’s thesis concludes that the multi-modality fusion method presented as feature fusion can be applied to the Mask R-CNN model to improve the networks performance by 1.8%points (1.8%pt.) bounding box mAP and 1.6%pt. segmentation mAP on SICK RGBD, 7.7%pt. bounding box mAP and 7.4%pt. segmentation mAP on ToF, and 7.4%pt. bounding box mAP and 7.4%pt. segmentation mAP on NYUDepth. The RTMDet model saw little to no improvements from the inclusion of depth but had similar baseline performance as the improved Mask R-CNN model that utilized feature fusion. The input extension method saw no improvements to performance as it faced technical implementation limitations.

No description available.

No description available.

No description available.