Recent advances of 3D acquisition devices have enabled large-scale acquisition of 3D scene data. Such data, if completely and well annotated, can serve as useful ingredients for a wide spectrum of computer vision and graphics works such as data-driven modeling and scene understanding, object detection and recognition. However, annotating a vast amount of 3D scene data remains challenging due to the lack of an effective tool and/or the complexity of 3D scenes (e.g. clutter, varying illumination conditions). This paper aims to build a robust annotation tool that effectively and conveniently enables the segmentation and annotation of massive 3D data. Our tool works by coupling 2D and 3D information via an interactive framework, through which users can provide high-level semantic annotation for objects. We have experimented our tool and found that a typical indoor scene could be well segmented and annotated in less than 30 minutes by using the tool, as opposed to a few hours if done manually. Along with the tool, we created a dataset of over a hundred 3D scenes associated with complete annotations using our tool. Both the tool and dataset will be available at http://scenenn.net.

Underwater monocular visual simultaneous localization and mapping (SLAM) plays a vital role in underwater computer vision and robotic perception fields. Unlike the autonomous driving or aerial environment, performing robust and accurate underwater monocular SLAM is tough and challenging due to the complex aquatic environment and the collected critically degraded image quality. The underwater images’ poor visibility, low contrast, and color distortion result in ineffective and insufficient feature matching, leading to the poor or even failure of the existing SLAM algorithms. To address this issue, we propose introducing the generative adversarial network (GAN) to perform effective underwater image enhancement before conducting SLAM. Considering the inherent real-time requirement of SLAM, we conduct knowledge distillation to achieve GAN compression to reduce the inference cost, while achieving high-fidelity underwater image enhancement and real-time inference. The real-time underwater image enhancement acts as the image pre-processing to build a robust and accurate underwater monocular SLAM system. With the introduction of real-time underwater image enhancement, we can significantly promote underwater SLAM performance. The proposed method is a generic framework, which could be extended to various SLAM systems and achieve various scales of performance gain.

Modern artificial intelligence offers a novel and transformative approach to creating digital art across diverse styles and modalities like images, videos and 3D data, unleashing the power of creativity and revolutionizing the way that we perceive and interact with visual content. This paper reports on recent advances in stylized 3D asset creation and manipulation with the expressive power of neural networks. We establish a taxonomy for neural stylization, considering crucial design choices such as scene representation, guidance data, optimization strategies, and output styles. Building on such taxonomy, our survey first revisits the background of neural stylization on 2D images, and then presents in-depth discussions on recent neural stylization methods for 3D data, accompanied by a benchmark evaluating selected mesh and neural field stylization methods. Based on the insights gained from the survey, we highlight the practical significance, open challenges, future research, and potential impacts of neural stylization, which facilitates researchers and practitioners to navigate the rapidly evolving landscape of 3D content creation using modern artificial intelligence.