We propose an efficient method to deal with the matching and registration problem found in cross-source point clouds captured by different types of sensors. This task is especially challenging due to the presence of density variation, scale difference, a large proportion of noise and outliers, missing data, and viewpoint variation. The proposed method has two stages: in the coarse matching stage, we use the ensemble of shape functions descriptor to select potential K regions from the candidate point clouds for the target. In the fine stage, we propose a scale embedded generative Gaussian mixture models registration method to refine the results from the coarse matching stage. Following the fine stage, both the best region and accurate camera pose relationships between the candidates and target are found. We conduct experiments in which we apply the method to two applications: one is 3D object detection and localization in street-view outdoor (LiDAR/VSFM) cross-source point clouds and the other is 3D scene matching and registration in indoor (KinectFusion/VSFM) cross-source point clouds. The experiment results show that the proposed method performs well when compared with the existing methods. It also shows that the proposed method is robust under various sensing techniques, such as LiDAR, Kinect, and RGB camera.

As digital watermarking has become an important tool for copyright protection, various watermarking schemes have been proposed in literature. Among them both discrete wavelet transform (DWT) and singular value decomposition (SVD) are commonly used. In a DWT-based watermarking scheme, the host image is decomposed into four frequency bands, and DWT coefficients in each band are modified to hide watermark information. Modification in all frequencies enables watermarking schemes using DWT robust to a wide range of attacks. However, as most transform methods, DWT decomposes images in terms of a standard basis set which is not necessarily optimal for a given image. By contrast with DWT, SVD offers a tailor-made basis for a given image which packs maximum signal energy into as few coefficients as possible. SVD is used in image processing also for its properties of stability, proportion invariance and rotation invariance. In this paper we propose a hybrid DWT-SVD domain watermarking scheme considering human visual properties. After decomposing the host image into four subbands, we apply SVD to each subband and embed singular values of the watermark into them. The embedding strength is determined by a human visual model proposed in A.S. Lewis and G. Knowles, (1992) and improved in M. Bertran et al., (2001). Our scheme has advantages of robustness for its embedding data into all frequencies and large capacity for using SVD. In addition, the use of human visual model guarantees the imperceptibility of the watermark.

2-D neighborhood preserving projection (2DNPP) uses 2-D images as feature input instead of 1-D vectors used by neighborhood preserving projection (NPP). 2DNPP requires less computation time than NPP. However, both NPP and 2DNPP use the L₂ norm as a metric, which is sensitive to noise in data. In this paper, we proposed a novel NPP method called low-rank 2DNPP (LR-2DNPP). This method divided the input data into a component part that encoded low-rank features, and an error part that ensured the noise was sparse. Then, a nearest neighbor graph was learned from the clean data using the same procedure as 2DNPP. To ensure that the features learned by LR-2DNPP were optimal for classification, we combined the structurally incoherent learning and low-rank learning with NPP to form a unified model called discriminative LR-2DNPP (DLR-2DNPP). By encoding the structural incoherence of the learned clean data, DLR-2DNPP could enhance the discriminative ability for feature extraction. Theoretical analyses on the convergence and computational complexity of LR-2DNPP and DLR-2DNPP were presented in details. We used seven public image databases to verify the performance of the proposed methods. The experimental results showed the effectiveness of our methods for robust image representation.

3D GAN inversion aims to achieve high reconstruction fidelity and reasonable 3D geometry simultaneously from a single image input. However, existing 3D GAN inversion methods rely on time-consuming optimization for each individual case. In this work, we introduce a novel encoder-based inversion framework based on EG3D, one of the most widely-used 3D GAN models. We leverage the inherent properties of EG3D’s latent space to design a discriminator and a background depth regularization. This enables us to train a geometry-aware encoder capable of converting the input image into corresponding latent code. Additionally, we explore the feature space of EG3D and develop an adaptive refinement stage that improves the representation ability of features in EG3D to enhance the recovery of fine-grained textural details. Finally, we propose an occlusion-aware fusion operation to prevent distortion in unobserved regions. Our method achieves impressive results comparable to optimization-based methods while operating up to 500 times faster. Our framework is well-suited for applications such as semantic editing.