https://www.um.edu.mt/library/oar/handle/123456789/83406 2026-04-21T02:10:45Z 2026-04-21T02:10:45Z https://www.um.edu.mt/library/oar/handle/123456789/83524 2021-11-09T05:40:23Z 2021-01-01T00:00:00Z

Title: The effect of orientation and pose on spatial relation detection Abstract: Detecting relationships between objects is an important way to thoroughly understand images. In this work we explore the effect of human pose, as a proxy for orientation, on detecting a specifi c subset of visual relationships, namely Spatial Relations. We use a human pose detector to detect 2D poses in images, impute the poses to correct missing joints and then encode the full poses in a number of representations. A number of models are trained using geometric and language features, incorporating pose features into a subset of the models. Overall, models trained without pose features produced better results. However, models trained with pose features showed a performance improvement for the relationship opposite. This signifi es the importance of pose features for the subset of Spatial Relations where the use of the intrinsic frame of reference is required. Description: M.Sc.(Melit.)

2021-01-01T00:00:00Z https://www.um.edu.mt/library/oar/handle/123456789/83523 2021-11-09T05:39:48Z 2021-01-01T00:00:00Z

Title: Mechanising symbolic controllability Abstract: Runtime Monitors are computational entities executed alongside programs to observe their behaviour with the aim of reaching a verdict about it. Since monitors are part of the Trusted Computing Base, they are expected to behave correctly; one property that is expected of them is deterministic behaviour. Nevertheless, this criterion is often specified in ambiguous terms. One candidate definition is consistent verdict detections [16]; this property was then characterised further through a symbolic analysis termed symbolic controllability to enable an automated analysis for monitors that involve data. This study investigates whether this symbolic analysis lends itself well for the implementation of a tool that determines whether a monitor exhibits this deterministic behaviour. Subsequently, we implement a tool that automates this symbolic analysis and evaluate it; we establish the worst-case complexity bounds of the designed algorithm and empirically evaluate the implementation. We also investigate multiple optimisation techniques for our implementation to improve its scalability and examine the potential gains incurred by these optimisations against a monitor benchmark. Description: M.Sc.(Melit.)

2021-01-01T00:00:00Z