With the advancement of AI-generated content technologies, AI-generated images (AGIs) have become increasingly influential in artistic creation and visual communication. However, the aesthetic quality of AGIs varies s...
详细信息
ISBN:
(数字)9798350368741
ISBN:
(纸本)9798350368758
With the advancement of AI-generated content technologies, AI-generated images (AGIs) have become increasingly influential in artistic creation and visual communication. However, the aesthetic quality of AGIs varies significantly due to technical limitations and the influence of user input, underscoring the urgent need for systematic aesthetic evaluation of AGIs. In addition, it is difficult to ensure the consistency of text-to-image, which compresses the application space of AGIs. To address these issues, a fine-grained dataset for Aesthetic and Alignment evaluation of AGIs (AGIAA-2K) is presented. This dataset contains 2,064 images generated using 172 well-designed prompts across six different AGI models, with each image annotated based on the subjective experiment. Then, the rationality of the dataset is verified by data analysis. Finally, the performances of the existing algorithms are evaluated in terms of image aesthetic assessment and text-to-image alignment of AGIs. The results demonstrate that these algorithms cannot effectively evaluate these two aspects. The AGIAA-2K is available at https://***/BoHu90/AGIAA-2K.
Edge extraction is a central problem in imageprocessing and it is a necessary step for computer vision tasks. In this paper, a novel global method P-GSG for edge extraction of image under Poisson noise is given, whic...
详细信息
This paper introduces Intermodule Data Bus (IMB)which is used for the C919 Flight Control System. It is necessary to develop IMB bus data processing software to provide data basis for civil aircraft airworthiness cert...
详细信息
In recent years, with the rapid development and popularization of artificial intelligence technology, CT has been more and more widely used in the field of clinical medicine as a non-volume micro body diagnostic metho...
详细信息
3D object detection is crucial for autonomous driving. Currently, most approaches depend on costly LiDAR sensors for acquiring precise depth information. Although pseudo-LiDAR techniques based on stereo cameras offer ...
详细信息
Due to the difficulty of acquiring a sufficient amount of satellite images in a short period of time, three-dimensional (3D) reconstruction with few-view satellite images has a wide range of application areas. However...
详细信息
ISBN:
(纸本)9798350379860;9798350379877
Due to the difficulty of acquiring a sufficient amount of satellite images in a short period of time, three-dimensional (3D) reconstruction with few-view satellite images has a wide range of application areas. However, based on few-view satellite images, most existing algorithms, including dense matching and the neural radiation fields, lose their stability when facing scenes with noticeable differences in elevation which is abbreviated as NDE-scenes. The greater the elevation difference, the more challenging the reconstruction becomes. To solve the problem of poor reconstruction in NDE-scenes, we propose a new framework based on neural radiation fields in this paper, along with two improvement methods: the signed distance function (SDF) field generation network (SDFGN) optimization strategy and the shadow prior module. It enables accurate and robust reconstruction of NDE-scenes with fewer views. Comparison and ablation experiments on the DFC2019 dataset verify the advancements of the proposed methods and the effectiveness of each module.
Recent development of clinical Computed tomography (CT) technologies has led to research for novel CT systems that allow safer and faster imaging, such as low-dose cardiac CT imaging via stationary CT. However, the co...
详细信息
image encryption using a chaotic system is a widely used technique. But, cryptanalysis of most chaotic encryption algorithms pointed out the disadvantages or drawbacks of the chaotic systems. As a remedy, a new and mo...
详细信息
Traditional multi-view target detection can only obtain the light intensity information of the target, and its detection ability is very limited in complex environments;Traditional polarization target detection mainly...
详细信息
With the increasing demand for autonomous vehicles, the reliable detection and recognition of objects have become paramount. Millimeter-wave radar systems operating at frequencies such as 94GHz offer several advantage...
详细信息
暂无评论