Point cloud completion is able to estimate the complete point cloud starting from the missing point cloud, which obtains higher quality point cloud data for widely used in remote sensing 3D modeling, medical imaging, robot vision, etc. The challenge of point clouds mainly lies in the disordered and unstructured nature, which makes point cloud completion difficult. Point cloud completion research can be broadly categorized into traditional approaches and deep learning based methods. Recently, intensive research has primarily focused on deep learning based methods, given robustness and efficiency in processing the substantial missing data encountered in complex real world scenes. In addition, deep learning based methods have higher generalization performance. To stimulate future research, this survey presents a comprehensive review of existing traditional and deep learning based 3D point cloud completion methods. This review conducts extensive examinations of each stage of the process, providing a compilation of famous datasets, metrics, and their respective characteristics. Additionally, the impacts of subsequent downstream application tasks with or without completion are discussed, followed by some potential future issues in point cloud completion

Semantic segmentation of large-scale point clouds is an indispensable component of outdoor scene perception, providing essential 3D semantic insights for applications in scene reconstruction, urban planning, autonomous driving, and more. However, the discriminative capability of point clouds features declines with increasing distance from the sensor, causing current methods to usually perform poorly in segmenting distant objects. To overcome this challenge and improve the differentiation between classes with similar geometric features, we propose the Cross-modal Semantic Focus Network (CSFNet). Firstly, we design a Multi-scale feature Dynamic Fusion module (MDF) to leverage multi-scale image features, thereby enriching the feature representation of point clouds with additional images color and texture information. Then, in order to extract the distinguishing features of distant and different categories of objects more efficiently, we propose a Semantic Focus Module (SFM) that employs a multi-class contrastive learning strategy to enhance feature discrimination. Finally, we introduce cross-modal knowledge distillation to augment the model’s comprehension of point clouds. Extensive experiments conducted on the SemanticKITTI and nuScenes datasets demonstrate the effectiveness of our method. Notably, our method achieves superior segmentation accuracy across multiple classes at various distances compared to current methods.

Semantic perception in driving scenarios plays a crucial role in intelligent transportation systems. However, existing Transformer-based semantic segmentation methods often do not fully exploit their potential in understanding driving scene dynamically. These methods typically lack spatial reasoning, failing to effectively correlate image pixels with their spatial positions, leading to attention drift. To address this issue, we propose a novel architecture, the Hierarchical Spatial Perception Transformer (HSPFormer), which integrates monocular depth estimation and semantic segmentation into a unified framework for the first time. We introduce the Spatial Depth Perception Auxiliary Network (SDPNet), a framework for multiscale feature extraction and multilayer depth map prediction to establish hierarchical spatial coherence. Additionally, we design the Hierarchical Pyramid Transformer Network (HPTNet), which uses depth estimation as learnable position embeddings to form spatially correlated semantic representations and generate global contextual information. Experiments on benchmark datasets such as KITTI-360, Cityscapes, and NYU Depth V2, demonstrate that HSPFormer outperforms several state-of-the-art networks, and achieves promising performance with 66.82% top-1 mIoU on KITTI-360, 83.8% mIoU on Cityscapes, and 57.7% mIoU on NYU Depth V2, respectively. The code will be made publicly available at https://github.com/SY-Ch/HSPFormer.