We study how small, fine-tuned generative large language models (LLMs) can moderate free-text search queries for surveillance video systems. Four open models, Llama 3.2 1B, Llama 3.2 3B, Qwen 2.5 0.5B, and 1.5 B, are trained on six subtasks: safety judgement, problem detection, target detection, span detection, safety explanation, and rephrase. The training combines a public toxicity set with abou

This thesis examines how gay male fans in mainland China understand their admiration for Taiwanese female Mandopop singers, and how they navigate the political controversies that surrounding Taiwanese female singers on Weibo. Drawing on qualitative data from eight semi-structured interviews, it applies a dual theoretical framework—Queer Cultural Decoding and Celebrity Politics under Digital Author

BACKGROUND: This systematic review aims to determine the loads of anti-A and anti-B in plasma-containing blood products that can safely be transfused to adults.MATERIALS AND METHODS: A comprehensive literature search was performed in PubMed, including studies dating back to 1940 until January 1 st, 2023. In total 1,869 articles were screened, describing 86 cases of hemolytic transfusion reactions

Radiolabeled targeting agents have emerged as valuable tools for the treatment of disseminated cancer. Monoclonal antibodies (mAbs) are widely employed as carriers for diagnostic and therapeutic radionuclides due to their exceptional specificity and affinity. However, their prolonged circulatory half-life can diminish diagnostic efficacy and increase radiation exposure to non-target tissues in the

In 2020, a case report described autologous transplantation of iPSC-derived dopamine (DA) neurons in a Parkinson's disease (PD) patient.1 The team now follows up with the pre-clinical safety and efficacy data of autologous iPSC-derived DA neurons, forming the basis for regulatory approval of a phase 1 clinical trial involving 8 patients.2

During the past decade, scholars have produced a vast amount of research using linked historical individual-level data, shaping and changing our understanding of the past. This linked data revolution has been powered by methodological and computational advances, partly focused on supervised machine-learning methods that rely on training data. The importance of obtaining high-quality training data

This thesis examines how innovations in electric aircraft and eVTOL vehicles can contribute to sustainable air travel and transform future transport systems. The research focuses on identifying the most promising technologies and how they could be applied in practice. The study is based on a systematic review of academic literature, case studies and expert interviews from both academia and the avi

This project explores how a beginner friendly music production device can lower the threshold for engaging with sound creation and rhythm making. Many existing tools rely on prior knowledge or technical skill, which discourages new users from experimenting in creative ways. This project instead focuses on quick feedback, clear interaction and a tactile experience that invites curiosity and play. T

Rapporten undersöker hur brusreducering påverkar ljudklassificering. Olika aspekter undersöks; informationsbevarande, hur bra brusreduceringen blir, samt brusnivåers och antalet klassers inverkan. Ljudklassificering är när en dator analyserar ljud och identifierar vilka klasser ljudet tillhör. För att skapa klassificering med maskininlärning, tränas ett nätverk att till bästa förmåga klassificerIt is often impossible to capture the sound one wants without capturing noise. Consequently, research has been done on noise filtering in audio signals. Currently deep learning algorithms are used for noise filtering. This thesis investigated the effects of noise and denoising on sound classification. The thesis used two standard datasets for clean sound and noise. FSD50K was chosen as clean soun

This thesis reviews existing research on how cybersecurity affects flight safety and to what extent these threats are addressed in regulatory requirements. This mixed method thesis employs anchoring in safety science, systematic literature review, legal requirements analysis, interviews, and a survey. Research shows risks related to unencrypted VHF/HF radio, CPDLC, ACARS, Mode-S, ADS-B, TCAS and G

Radar has come a long way since its invention in the 1930s. It is not only used to detect the distance and angular position of a target but can also measure velocity thanks to the Doppler frequency shift. In more advanced systems, the radar can perform tasks such as imaging and classification of the target. In some cases, full or partial polarimetry is used to gain additional information about the

This study is written for the purpose to analyze how automation in today’s modern flight deck is affecting pilots basic flying skills. It is focused on the pilot’s ability to control thrust. The main thesis transpires that pilots may had a better physical understanding around aerodynamics and speed control during basic flying training, operating aircraft with minimal or no automation (energy manag

Abstract: This bachelor thesis explores the concept of just culture within aviation, with a specific focus on European airline captains’ knowledge, understanding, and perception of their reporting obligations. Grounded in the theoretical just culture framework of safety culture, developed by James Reason (1997), the study examines whether the principles of just culture - now nearly 30 years ol

This thesis aims to identify current applications of artificial intelligence in airline pilot training, review existing research, evaluate potential improvements, and examine associated limitations. A literary study was conducted to achieve this. Open research was found to be limited but broad, showing a wide range of potential uses. A few airlines and training providers were found to be currently