The increasing demands on interactive algorithm visualisations that can be effectively used in learning, teaching and research, as well as the lack of up-to-date visualisation systems calls for a new modern architectu...
详细信息
Face morphing attacks target to circumvent Face Recognition Systems (FRS) by employing face images derived from multiple data subjects (e.g., accomplices and malicious actors). Morphed images can be verified against c...
详细信息
In summary, the ROSE system has been designed to allow efficiency to be combined with fiexibility for support of interactive computergraphics applications. The object paradigm is exploited throughout all aspects of t...
详细信息
Incremental subspace methods have proven to enable efficient training if large amounts of training data have to be processed or if not all data is available in advance. In this paper we focus on incremental LDA learni...
ISBN:
(纸本)1901725340
Incremental subspace methods have proven to enable efficient training if large amounts of training data have to be processed or if not all data is available in advance. In this paper we focus on incremental LDA learning which provides good classification results while it assures a compact data representation. In contrast to existing incremental LDA methods we additionally consider reconstructive information when incrementally building the LDA subspace. Hence, we get a more flexible representation that is capable to adapt to new data. Moreover, this allows to add new instances to existing classes as well as to add new classes. The experimental results show that the proposed approach outperforms other incremental LDA methods even approaching classification results obtained by batch learning.
Measuring the similarity between 3D models is a fundamental task in 3D inodels retrieval. In this paper, we propose a new method based on Global Geometric Feature Map (GGFM) to represent arbitrary polygonal 3D models....
详细信息
ISBN:
(纸本)1424400287
Measuring the similarity between 3D models is a fundamental task in 3D inodels retrieval. In this paper, we propose a new method based on Global Geometric Feature Map (GGFM) to represent arbitrary polygonal 3D models. Since 3D polygonal model can be expressed as a set of facets, the GGFM contrast constitute a spherical histogram about the normal orientation and area and position of every facet on the surface of the model. By computing the spherical correlation between the GGFMs of the matched models, similarity of two models can be obtained. Experimental results show that the proposed method performs well in 3D model similarity matching and is invariant to the translation and rotation and scaling of 3D model. Comparing to the existing methods, this method is fast and needs low computation and storage cost since each facet of the model needs to be computed only once in GGFM.
Face recognition (FR) systems continue to spread in our daily lives with an increasing demand for higher explainability and interpretability of FR systems that are mainly based on deep learning. While bias across demo...
详细信息
Over the past years, the main research innovations in face recognition focused on training deep neural networks on large-scale identity-labeled datasets using variations of multi-class classification losses. However, ...
详细信息
Let P (k, n) be the set of products of k factors from the set {1, . . ., n}. In 1955, Erdős posed the problem of determining the order of magnitude of |P (2, n)| and proved that |P (2, n)| = o(n2) for n → ∞. In 201...
详细信息
With fast 3D graphics becoming more and more available even on low end platforms, the focus in hardware-accelerated rendering is beginning to shift towards higher quality rendering and additional functionality instead...
详细信息
ISBN:
(纸本)0201485605
With fast 3D graphics becoming more and more available even on low end platforms, the focus in hardware-accelerated rendering is beginning to shift towards higher quality rendering and additional functionality instead of simply higher performance implementations based on the traditional graphics pipeline. In this paper we present techniques for realistic shading and lighting using computergraphics hardware. In particular, we discuss multipass methods for high quality local illumination using physically-based reflection models, as well as techniques for the interactive visualization of non-diffuse global illumination solutions. These results are then combined with normal mapping for increasing the visual complexity of rendered images. Although the techniques presented in this paper work at interactive frame rates on contemporary graphics hardware, we also discuss some modifications of the rendering pipeline that help to further improve both performance and quality of the proposed methods. Copyright ACM 1999.
Recently, significant progress has been made in face presentation attack detection (PAD), which aims to secure face recognition systems against presentation attacks, owing to the availability of several face PAD datas...
详细信息
暂无评论