OBJECTIVE: Early diagnosis of Alzheimer's disease depends on accessible cognitive assessments, such as the Rey-Osterrieth Complex Figure (ROCF) test. However, manual scoring of this test is labor-intensive and subjective, which introduces experimental biases. Additionally, deep learning models face challenges due to the limited availability of annotated clinical data, particularly for assessments like the ROCF test. This scarcity of data restricts model generalization and exacerbates domain shifts across different populations. METHODS: We propose a novel framework comprising a data synthesis pipeline and ROCF-Net, a deep learning model specifically designed for ROCF scoring. The synthesis pipeline is lightweight and capable of generating realistic, diverse, and annotated ROCF drawings. ROCF-Net, on the other hand, is a cross-domain scoring model engineered to address domain discrepancies in stroke texture and line artifacts. It maintains high scoring accuracy through a novel line-specific attention mechanism tailored to the unique characteristics of ROCF drawings. RESULTS: Unlike conventional synthetic medical imaging methods, our approach generates ROCF drawings that accurately reflect Alzheimer's-specific abnormalities with minimal computational cost. Our scoring model achieves SOTA performance across differently sourced datasets, with a Mean Absolute Error (MAE) of 3.53 and a Pearson Correlation Coefficient (PCC) of 0.86. This demonstrates both high predictive accuracy and computational efficiency, outperforming existing ROCF scoring methods that rely on Convolutional Neural Networks (CNNs) while avoiding the overhead of parameter-heavy transformer models. We also show that training on our synthetic data generalizes as well as training on real clinical data, where the difference in performance was minimal (MAE differed by 1.43 and PCC by 0.07), indicating no statistically significant performance gap. CONCLUSION: Our work introduces four contributions: (1) a cost-effective pipeline for generating synthetic ROCF data, reducing dependency on clinical datasets; (2) a domain-agnostic model for automated ROCF scoring across diverse drawing styles; (3) a lightweight attention mechanism aligning model decisions with clinical scoring for transparency; and (4) a bias-aware framework using synthetic data to reduce demographic disparities, promoting fair cognitive assessment across populations.

Abstract Achieving photorealism when rendering virtual scenes in movies or architecture visualizations often depends on providing a realistic illumination and background. Typically, spherical environment maps serve both as a natural light source from the Sun and the sky, and as a background with clouds and a horizon. In practice, the input is either a static high‐resolution HDR photograph manually captured on location in real conditions, or an analytical clear sky model that is dynamic, but cannot model clouds. Our approach bridges these two limited paradigms: a user can control the sun position and cloud coverage ratio, and generate a realistically looking environment map for these conditions. It is a hybrid data‐driven analytical model based on a modified state‐of‐the‐art GAN architecture, which is trained on matching pairs of physically‐accurate clear sky radiance and HDR fisheye photographs of clouds. We demonstrate our results on renders of outdoor scenes under varying time, date and cloud covers. Our source code and a dataset of 39 000 HDR sky images are publicly available at https://github.com/CGGMFF/SkyGAN .

The topic of visual saliency is well-known and spreads across numerous disciplines. In this paper, we examine how saliency models perform on images with specific characteristics. We explore a saliency of 14 different digitized paintings, all representing a biblical scene of The Last Supper. We evaluate the performance of different saliency models. The models are using the traditional approach as well as the deep learning approach. For the evaluation, we use three different metrics, AUC, NSS and CC. As ground truth, we use eye-tracking data from 35 participants. Our analysis shows that deep-learning methods predict the most salient parts of the paintings very similar to real eye fixations. We also checked the consistency of gaze patterns among the participants.

Classification of 3D objects - the selection of a category in which each object belongs - is of great interest in the field of machine learning. Numerous researchers use deep neural networks to address this problem, altering the network architecture and representation of the 3D shape used as an input. To investigate the effectiveness of their approaches, we conduct an extensive survey of existing methods and identify common ideas by which we categorize them into a taxonomy. Second, we evaluate 11 selected classification networks on two 3D object datasets, extending the evaluation to a larger dataset on which most of the selected approaches have not been tested yet. For this, we provide a framework for converting shapes from common 3D mesh formats into formats native to each network, and for training and evaluating different classification approaches on this data. Despite being partially unable to reach the accuracies reported in the original papers, we compare the relative performance of the approaches as well as their performance when changing datasets as the only variable to provide valuable insights into performance on different kinds of data. We make our code available to simplify running training experiments with multiple neural networks with different prerequisites.

In this paper we present a new technique for the display of High Dynamic Range (HDR) images on Low Dynamic Range (LDR) displays. The described process has three stages. First, the input image is segmented into luminance zones. Second, the tone mapping operator (TMO) that performs better in each zone is automatically selected. Finally, the resulting tone mapping (TM) outputs for each zone are merged, generating the final LDR output image. To establish the TMO that performs better in each luminance zone we conducted a preliminary psychophysical experiment using a set of HDR images and six different TMOs. We validated our composite technique on several (new) HDR images and conducted a further psychophysical experiment, using an HDR display as reference, that establishes the advantages of our hybrid three-stage approach over a traditional individual TMO.

A major obstacle for real-time rendering of high-fidelity graphics is computational complexity. A key point to consider in the pursuit of “realism in real time” in computer graphics is that the Human Visual System (HVS) is a fundamental part of the rendering pipeline. The human eye is only capable of sensing image detail in a 2 ˆ foveal region, relying on rapid eye movements, or saccades, to jump between points of interest. These points of interest are prioritized based on the saliency of the objects in the scene or the task the user is performing. Such “glimpses” of a scene are then assembled by the HVS into a coherent, but inevitably imperfect, visual perception of the environment. In this process, much detail, that the HVS deems unimportant, may literally go unnoticed. Visual science research has identified that movement in the background of a scene may substantially influence how subjects perceive foreground objects. Furthermore, recent computer graphics work has shown that both fixed viewpoint and dynamic scenes can be selectively rendered without any perceptual loss of quality, in a significantly reduced time, by exploiting knowledge of any high-saliency movement that may be present. A high-saliency movement can be generated in a scene if an otherwise static objects starts moving. In this article, we investigate, through psychophysical experiments, including eye-tracking, the perception of rendering quality in dynamic complex scenes based on the introduction of a moving object in a scene. Two types of object movement are investigated: (i) rotation in place and (ii) rotation combined with translation. These were chosen as the simplest movement types. Future studies may include movement with varied acceleration. The object's geometry and location in the scene are not salient. We then use this information to guide our high-fidelity selective renderer to produce perceptually high-quality images at significantly reduced computation times. We also show how these results can have important implications for virtual environment and computer games applications.

Typically, content-based image retrieval (CBIR) systems receive an image or an image description as input and retrieve images from a database that are similar to the query image in regard to properties such as color, texture, shape, or layout. A kind of system that did not receive much attention compared to CBIR systems, is one that searches for images that are not similar but exact copies of the same image that have undergone some transformation. In this paper, we present such a system referred to as an image fingerprinting system, since it aims to extract unique and robust image descriptors (in analogy to human fingerprints). We examine the use of color-based descriptors and provide comparisons for different quantization methods, histograms calculated using color-only and/or spatial-color information with different similarity measures. The system was evaluated with receiver operating characteristic (ROC) analysis on a large database of 919 original images consisting of randomly drawn art images and similar images from specific categories, along with 30 transformed images for each original, totaling 27570 images. The transformed images were produced with attacks that typically occur during digital image distribution, including different degrees of scaling, rotation, cropping, smoothing, additive noise and compression, as well as illumination contrast changes. Results showed a sensitivity of 96% at the small false positive fraction of 4% and a reduced sensitivity of 88% when 13% of all transformations involved changing the illuminance of the images. The overall performance of the system is encouraging for the use of color, and particularly spatial chromatic descriptors for image fingerprinting