This paper explores the challenges of converting architectural floor plans from raster to vector images. Unlike previous studies, our research focuses on domain adaptation to address stylistic and technical variations across different floor plan datasets. We develop and test our vectorization method on the CubiCasa5K benchmark, which includes 3 different floor plan styles. Our analysis reveals differences in input features across the CubiCasa5K styles, indicating the potential for domain adaptation research, mostly in room segmentation. However, we also find multiple indications that labelling in the CubiCasa5K dataset is ambiguous and inconsistent. Furthermore, styles with more training data do not always perform better, highlighting the complexity differences between floor plan styles. Our baseline shows a 0.7% gap for rooms yet a 0.6% improvement for objects, likely caused by the smaller feature gaps and inconsistent labelling. To address the adaptation gap, we add a Multi-Kernel Maximum Mean Discrepancy (MK-MMD) loss to the CubiCasa5K model to minimize feature distribution differences between domains. While our MK-MMD implementation shows potential for reducing the adaptation gap, persistence issues and mixed results across classes make it difficult to draw clear conclusions. Our findings also show the role of balancing spatial context in the MK-MMD calculation. These insights lay a foundation for future domain adaptation research in floor plan vectorization.

Tiny object detection (TOD) in satellite imagery is critical for applications including pipeline monitoring, where the detection of tiny objects, such as excavators near the pipeline networks, can prevent potential incidents. However, TOD faces challenges due to the limited pixel representation of objects, complications with Intersection over Union-based label assignment, and semantic confusion between visually similar classes. This paper introduces OE-Net, a modular extension of Oriented R-CNN,
specifically designed for TOD in satellite images. OE-Net integrates three novel components: a Dilated Fusion module to enhance the feature extraction, a Balanced Ranking Assigner to improve the anchor matching, and Vector Mapping for more precise classification of visually similar objects. The modular design of OE-Net allows its novel components to be easily integrated into other detection models. Tested on the TinyDOTA dataset, OE-Net sets a new benchmark in TOD performance, achieving a 6.3
percentage points improvement in mean Average Precision (mAP) over the Oriented R-CNN baseline. On a novel excavator detection dataset, named ExcaSat and developed for pipeline monitoring, OE-Net outperforms the Oriented R-CNN baseline by 11.1 percentage points in mAP. Furthermore, OE-Net surpasses state-of-the-art methods on both TinyDOTA and ExcaSat, establishing a new standard in computational efficiency and detection precision for satellite-based monitoring tasks. The code is available at: https://github.com/OE-jimvanoosten/mmrotate-OE.git.

The realm of architectural design and conceptualization, despite witnessing advancements in design complexity facilitated by technological tools and fabrication techniques, appears to have experienced limited transformative change over time in the way we begin our design process. But is there an alternative way to look at architecture ? Perhaps a completely different way to begin a project ? Can the age of data abundance cause a shift in the way we look at architectural design ?

While the performance of traditional confidence-based rejectors is heavily dependent on the calibration of the pretrained model, this study proposes the concept of feature-based rejectors and the whole pipeline where such rejector can be used in. Multiple design and development decisions along the implementation are discussed in the paper - involving humans in the loop of creating the feature set that machine should use and defining value, using interpretability methods such as saliency map to extract the features that the machine actually uses, and SceneRejector is the end product of the whole process. It is relevant to the field of computer vision, and it is applied to the task of scene classification within the scope of this paper. Its performance is evaluated against baselines with accuracy, rejection rate, and a new metric named value, which more comprehensively measures the practical value of machine learning with a reject option in different use cases. Experiments have proved the concept of a feature-based rejector works and it is able to filter out unknown unknowns, which is a challenge to the confidence-based rejectors. It also creates better value than the traditional confidence-based rejectors in some cases, especially when the machine learning model is not well calibrated. Further analysis is conducted to understand the behavior of SceneRejector as well. It is discovered that SceneRejector brings better value than confidence-based rejectors when the pretrained model is not well calibrated, and that rejection rate and accuracy of the rejector is also correlated with the value.

There is growing research on automated video summarization following the rise of video content. However, the subjectivity of the task itself is still an issue to address. This subjectivity stems from the fact that there can be different summaries for the same video depending on which parts one considers important. Supervised models especially suffer from this problem as they need informative labels to learn from. As a result, upon evaluation, supervised models appear to perform worse than unsupervised models. This inspired our research on whether action localization can aid the video summarization process. To investigate this issue, this paper will answer the question of how well VASNet, a supervised video summarization model, can predict summaries for videos in an action localization dataset. This involves investigating whether action localization can produce well-correlated human-generated summaries and how it affects the quality of predicted summaries. Our findings reveal that there is a positive indication that action localization can aid in producing more well-correlated human summaries. In addition, we have observed that upon comparison with several video summarization models, VASNet has performed well and that in general, supervised models appear to outperform unsupervised ones when trained with an action localization dataset.

In this paper, the DSNet framework used for automatic video summarization gets reviewed when using action localization datasets. The problem facing video summarizations using deep learning techniques is that datasets can be subjective depending on preferences of human annotators, making for noise in the labeling. This paper will look at a anchor-based approach and anchor-free approach which were introduced by the DSNet framework. More specific it will evaluate in experiments using different hyper-parameters if these approaches gain an increased performances when using action localization datasets instead. These results will show the increase in accuracy when using action localization datasets. Moreover it will compare the different approaches, meaning anchor-based and anchor-free, and see if they still have comparable performance with the method.

In the problem of video summarization, the goal is to select a subset of the input frames conveying the most important information of the input video. The collection of data proves to be a challenging task. In part because there exists a disagreement among human annotators on what segments of a video should be considered important for a summary. In this study we analyse a new dataset created with the goal of increasing agreement between the human annotators. The dataset has been created with the use of a novel annotation method, which uses existing action localization labels for segmenting the videos. We train a supervised and an unsupervised deep learning framework on popularly used benchmark datasets and the new dataset. Experimental results show the effectiveness of this novel summary annotation method in improving
the agreement between annotators. Analysis reveals some issues with the evaluation of the deep learning framework.

Video summarization is a task which many researchers have tried to automate with deep learning methods. One of these methods is the SUM-GAN-AAE algorithm developed by Apostolidis et al. which is an unsupervised machine learning method evaluated in this study. The research aims at testing the algorithm's performance on the Breakfast dataset, which is an action localization dataset, and evaluate it with rank correlation coefficients. Parameter optimization was performed to tune the learning rate of the system according to the Breakfast dataset. Then, by using k-fold cross-validation, three metrics were used to evaluate the trained model - F-Score, Kendall's τ and Spearman's ρ. Analysis of the results indicates a high F-Score as reported by the SUM-GAN-AAE paper but low rank correlation coefficients. Moreover, plotting importance scores per frame demonstrates the algorithm's inability to select key frames. The findings suggest that F-Score is not a fitting metric to use in the context of video summarization and the SUM-GAN-AAE algorithm performs poorly not only on action localization datasets but also on video summarization ones such as SumMe.