Browsing by Author "Jacob, Christian J"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    3D Design Review Systems in Immersive Environments
    (2023-09-21) Addam, Omar Khodr; Maurer, Frank O; Willett, Wesley; Jacob, Christian J; El-Hacha, Raafat; Irani, Pourang Polad
    Design reviews play a crucial role in the development process, ensuring the quality and effectiveness of designs in various industries. However, traditional design review methods face challenges in effectively understanding and communicating complex 3D models. Immersive technologies, particularly Head-Mounted Displays (HMDs), offer new opportunities to enhance the design review process. In this thesis, we investigate using immersive environments, specifically HMDs, for 3D design reviews. We begin with a systematic literature review to understand the current state of employing HMDs in industry for design reviews. As part of this review, we utilize a detailed taxonomy from the literature to categorize and analyze existing approaches. Additionally, we present four iterations of an immersive design review system developed during my industry experience. Two of these iterations are evaluated through case studies involving domain experts, including engineers, designers, and clients. A formal semi-structured focus group is conducted to gain further insights into traditional design review practices. The outcomes of these evaluations and the focus group discussions are thoroughly discussed. Based on the literature review and the focus group findings, we uncover a new challenge associated with using HMDs in immersive design reviews—asynchronous and remote collaboration. Unlike traditional design reviews, where participants view the same section on a shared screen, HMDs allow independent exploration of areas of interest, leading to a shift from synchronous to asynchronous communication. Consequently, important feedback may be missed as the lead designer disconnects from the users' perspectives. To address this challenge, we collaborate with a domain expert to develop a prototype that utilizes heatmap visualization to display 3D gaze data distribution. This prototype enables lead designers to quickly identify areas of review and missed regions. The study incorporates the Design Critique approach and provides valuable insights into different heatmap visualization variants (top view projection, object-based, and volume-based). Furthermore, a list of well-defined requirements is outlined for future spatio-temporal visualization applications aimed at integrating into existing workflows. Overall, this thesis contributes to the understanding and improvement of immersive design review systems, particularly in the context of utilizing HMDs. It offers insights into the current state of employing HMDs for design reviews, utilizes a taxonomy from the literature to analyze existing approaches, highlights challenges associated with asynchronous collaboration, and proposes a prototype solution with heatmap visualization to address the identified challenge.
  • Thumbnail Image
    ItemOpen Access
    Quality-Based Face Recognition System
    (2017-12-20) Zohra, Fatema Tuz; Gavrilova, Marina; Jacob, Christian J; Dimitrov, Vassil Simeonov
    Quality assessment of a biometric sample is relatively difficult and understudied problem compared to the automated recognition and feature extraction in biometrics. More attention should be directed towards this problem since it has been found in many studies that the quality of samples significantly affects the performance of a biometric system. This thesis focuses on designing a unified framework which can adaptively compensate for different quality degradations of the facial images. The proposed quality estimation model determines the overall quality of a facial sample by considering the impact of quality degradation on the performance of the sample. Our proposed quality-based face recognition system utilizes this overall quality score to determine the appropriate preprocessing steps and facial representations for improved recognition performance. The proposed methodology employs a quality-based weighted score fusion to boost the recognition performance further. Extensive experiments with real and synthetic samples demonstrate the effectiveness of the proposed methodology.