image segmentation is essential in digital imageprocessing applications. Multilevel thresholding is a popular technique for image segmentation. An image can be divide into multiple classes. In this paper, the image q...
详细信息
Watermarks in historical manuscripts are figural shapes serving as tokens for provenance research (e.g. scribe identification, dating, papermill attribution, scribe-papermaker relation, trading, etc.) in Humanities su...
详细信息
ISBN:
(纸本)9783031705427;9783031705434
Watermarks in historical manuscripts are figural shapes serving as tokens for provenance research (e.g. scribe identification, dating, papermill attribution, scribe-papermaker relation, trading, etc.) in Humanities such as Musicology. As of today, they come in a variety of formats: digitized handtracings and rubbings, X-ray based imagery and, more recently, thermograms acquired with infrared (IR) cameras - all of which have been made accessible via image data bases in libraries or archives like the watermark information system (WZIS). A key use case from a scholar's perspective is the search for similar or even equal watermarks in whatever digitized data collections. Non-surprisingly, the prerequisite is the availability of a versatile, reliable, and user-friendly tool comprising methods from digital imageprocessing (IP) and pattern recognition (PR). In our paper, we focus on bridging the gap between digitized thermograms of music manuscripts and watermark classification for similarity-based search through (i) a state-of-the-art (SOTA) analysis, (ii) a resulting conceptual design based on well-understood SOTA as well as novel methods, (iii) an easy-to-use implementation, and (iv) an experimental validation as Proof-of-Concept (PoC). The current system performance is characterized using thermograms recently made openly available within the DRACMarkS project as well as WZIS. The experimental results clearly demonstrate success in bridging the existing gap hence also setting a baseline for an as yet lacking benchmark.
This study addresses the critical need for efficient quality control in textile manufacturing, aiming to enhance the production of high-quality fabric to meet consumer demands and industry benchmarks. Despite the meti...
详细信息
Invisible watermarks safeguard images' copyrights by embedding hidden messages only detectable by owners. They also prevent people from misusing images, especially those generated by AI models. We propose a family...
Finding test-cases that cause mission-critical behavior is crucial to increase the robustness of satellite on-board imageprocessing. Using genetic algorithms, we are able to automatically search for test cases that p...
详细信息
Although AI systems have been applied in various fields and achieved impressive performance, their safety and reliability are still a big concern. This is especially important for safety-critical tasks. One shared cha...
详细信息
ISBN:
(纸本)9783031282409;9783031282416
Although AI systems have been applied in various fields and achieved impressive performance, their safety and reliability are still a big concern. This is especially important for safety-critical tasks. One shared characteristic of these critical tasks is their risk sensitivity, where small mistakes can cause big consequences and even endanger life. There are several factors that could be guidelines for the successful deployment of AI systems in sensitive tasks: (i) failure detection and out-ofdistribution (OOD) detection;(ii) overfitting identification;(iii) uncertainty quantification for predictions;(iv) robustness to data perturbations. These factors are also challenges of current AI systems, which are major blocks for building safe and reliable AI. Specifically, the current AI algorithms are unable to identify common causes for failure detection. Furthermore, additional techniques are required to quantify the quality of predictions. All these contribute to inaccurate uncertainty quantification, which lowers trust in predictions. Hence obtaining accurate model uncertainty quantification and its further improvement are challenging. To address these issues, many techniques have been proposed, such as regularization methods and learning strategies. As vision and language are the most typical data type and have many open source benchmark datasets, this thesis will focus on vision-language data processing for tasks like classification, image captioning, and vision question answering. In this thesis, we aim to build a safeguard by further developing current techniques to ensure the accurate model uncertainty for safety-critical tasks.
There is a severe demand for, and shortage of, large accurately labeled datasets to train supervised computational intelligence (CI) algorithms in domains like unmanned aerial systems (UAS) and autonomous vehicles. Th...
详细信息
ISBN:
(纸本)9798350332285
There is a severe demand for, and shortage of, large accurately labeled datasets to train supervised computational intelligence (CI) algorithms in domains like unmanned aerial systems (UAS) and autonomous vehicles. This has hindered our ability to develop and deploy various computer vision algorithms in/across environments and niche domains for tasks like detection, localization, and tracking. Herein, we propose a new human-in-the-loop (HITL) based growing neural gas (GNG) algorithm to minimize human intervention during labeling large UAS data collections over a shared geospatial area. Specifically, we address human driven events like new class identification and mistake correction. We also address algorithm-centric operations like new pattern discovery and self-supervised labeling. The effectiveness of our algorithm is demonstrated using simulated realistic ray-traced low altitude UAS data from the Unreal Engine. Our results show that it is possible to increase speed and reduce mental fatigue over hand labeling large image datasets.
Detecting dengue fever using imageprocessing techniques typically involves the analysis of medical images such as blood smears or tissue samples. Dengue is a viral disease transmitted by Aedes mosquitoes, and its dia...
详细信息
Environment analysis is a critical part of autonomous vehicle for transport applications and for passenger safety. The solutions demonstrating the greatest robustness have been integrating multiple sensors used for re...
详细信息
The research presents a hybrid approach to identify and categorise nutritional deficiency syndrome in citrus leaves using imageprocessing and machine learning. The method includes processingimages, segmenting images...
详细信息
暂无评论