The neighborhood effect, a pivotal element within the realm of Geographic Cellular Automata (GCA) modeling, has garnered significant attention in research. However, no research has yet investigated GCA modeling based on varying neighborhood sensitivity for different land use types. In this study, we sought to bridge this gap by integrating the First Law of Geography with diverse sensitivities of different land use types, thus introducing a novel approach termed Adaptive Spatially Heterogeneous Neighborhood (ASHN) for GCA modeling. By applying this innovative framework to three regions, namely Beijing, Wuhan, and the Pearl River Delta, we elucidated the implementation process and conducted comprehensive land use change simulations. The calibration period spanned from 2000 to 2010, followed by the validation period from 2010 to 2020. The results demonstrated that the ASHN-GCA model outperformed both the Adaptive Homogeneous Neighborhood Geographic Cellular Automata (AHN-GCA) model and the Homogeneous Neighborhood Geographic Cellular Automata (HN-GCA) model, yielding superior Overall Accuracy (OA), kappa, fuzzy kappa, and Figure of Merit (FoM) scores. Furthermore, the ASHN-GCA model provided more nuanced and detailed insights into landscape patterns, further highlighting its efficacy and potential for advancing GCA modeling in land use dynamics.

The data-hungry nature of deep learning and the high cost of annotating point-level labels for point clouds make it difficult to apply semantic segmentation methods to unlabeled real-world indoor scenes. Therefore, label-efficient point cloud segmentation has become a promising research topic. We noticed that the online housing design platforms can provide a large number of synthetic indoor 3D scenes, which are created with semantic labels. In this paper, we propose to learn semantic segmentation on synthetic point clouds and adapt the model for unlabeled real-world data. The main challenge is that directly using models trained on synthetic data for real-world data produces poor results due to the large domain gap between synthetic and real-world data. We design a point cloud style transfer network and a feature discrimination network to reduce the domain gap in both the input space and the feature space. Experiments show that our approach significantly improves the performance on real-world data for models learned from synthetic data.

3D object reconstruction from a single image is a highly ill-posed problem, requiring strong prior knowledge of 3D shapes. Deep learning methods are popular for this task. Especially, most works utilized 3D deconvolution to generate 3D shapes. However, the resolution of results is limited by the high resource consumption of 3D deconvolution. In this paper, we propose SliceNet, sequentially generating 2D slices of 3D shapes with shared 2D deconvolution parameters. To capture relations between slices, the RNN is also introduced. Our model has three main advantages: First, the introduction of RNN allows the CNN to focus more on local geometry details,improving the results’ fine-grained plausibility. Second, replacing 3D deconvolution with 2D deconvolution reducs much consumption of memory, enabling higher resolution of final results. Third, an slice-aware attention mechanism is designed to provide dynamic information for each slice’s generation, which helps modeling the difference between multiple slices, making the learning process easier. Experiments on both synthesized data and real data illustrate the effectiveness of our method.