Human Motion Recovery (HMR) research mainly focuses on ground-based motions such as running. The study on capturing climbing motion, an off-ground motion, is sparse. This is partly due to the limited availability of climbing motion datasets, especially large-scale and challenging 3D labeled datasets. To address the insufficiency of climbing motion datasets, we collect AscendMotion, a large-scale well-annotated, and challenging climbing motion dataset. It consists of 412k RGB, LiDAR frames, and IMU measurements, including the challenging climbing motions of 22 skilled climbing coaches across 12 different rock walls. Capturing the climbing motions is challenging as it requires precise recovery of not only the complex pose but also the global position of climbers. Although multiple global HMR methods have been proposed, they cannot faithfully capture climbing motions. To address the limitations of HMR methods for climbing, we propose Climbing-Cap, a motion recovery method that reconstructs continuous 3D human climbing motion in a global coordinate system. One key insight is to use the RGB and LiDAR modalities to separately reconstruct motions in camera coordinates and global coordinates and to optimize them jointly. We demonstrate the quality of the AscendMotion dataset and present promising results from ClimbingCap. The AscendMotion dataset and source code release publicly at http://www.lidarhumanmotion.net/climbingcap/

This paper presents a Blur Resist Generative Adversarial Network (GAN) (BrGAN) with multiple joint dilated residual convolutions for chlorophyll image restoration of the Geostationary Ocean Color Imager (GOCI). First, a publicly available dataset was built to support this study. Second, a multiple attention perception mechanism and a multiple joint dilated residual convolution module was proposed to cope with the challenge of large missing areas in GOCI chlorophyll images. Third, a patch GAN based discrimination module was proposed to avoid the restored areas with generating mosaic and shadows. Our experimental results demonstrate that the BrGAN can reach 37.06 in the peak signal-to-noise ratio (PSNR) and 0.0485 in the Learned Perceptual Image Patch Similarity (LPIPS), respectively. The comparative study shows that the BrGAN achieves the highest effectiveness and advancement among other seven state-of-the-art methods.

Visualization in an online multiple user, multiple location cooperative working system for complicated 3D structures such as in architecture design is a great challenge. There are special requirements for such cooperative environment. Some of them are related to the need of multiple participants while the others are for the specific needs of the application. The paper presents the visualization scheme and its implementation for the cooperative 3D design system particularly for the application of architecture design for building construction.

To perform 3D editing in a cooperative environment presents problems that do not exist in other environments. Maintaining the consistency among all the replicated instances of the scene being edited is the key problem. The authors present the mechanisms to solve these problems such as collaborative operations on multiple scenes, cooperative copy and paste operations, and the ability of reversing actions (undo) in a cooperative editing environment. A brief introduction to a cooperative 3D editing and visualisation tool is presented. The protocol to ensure its memory consistency is discussed. The data structures, as well as the techniques used to manipulate the data to realise the mechanisms are described.

The paper presents a prototype system for cooperative architecture design. The system includes a multi-site cooperative 3D editor and a cooperative and communication support platform. By using the prototype system, the users are able to load a design scene, modify its properties and save the result cooperatively. The online basic cooperative, interactive editing functionality on the designed objects via the communication network has been achieved.

The paper describes a cooperative design system for a 3D virtual scene. The objective of the system is to allow multiple users to design a 3D scene cooperatively via a communication network. The system is based on a distributed layered architecture. At the application layer, a 3D cooperative design tool is implemented. The basic cooperative design operations include reading and saving a 3D scene, selecting an object from the common 3D scene, modification of the objects by geometrical transformations, modification of the attributes of the global scene or an individual object, such as, object texture, material, illumination parameters, etc. A special protocol (SMI) has been designed and implemented for the communication with the peer cooperative design tool situated on peer hosts. The session control layer and group communication layer form the cooperative support platform for the system. The session control layer is responsible for the tasks required by a cooperative working session such as service multiplexing, data consistency control, member admission etc. The group communication layer provides the point to point or point to multipoint communication. The experimental system is currently implemented on Silicon Graphics hardware platform with a 100 MB/s internal network. Open Inventor has been used for the 3D design tool. The data format is fully VRML compatible.

There is a lot of development currently being carried out on cooperative visualization via communication networks. User interaction is kept to minimum in this kind of development, such as only making marks on the transmitted images, etc. There is relatively less work being done on complicated cooperative interactions, especially real-time cooperative interaction via networks. The work presented in this paper is a challenge to this type of cooperative work. The objective of the system is to allow multiple users to design a 3D scene cooperatively via communication networks in real time. The paper describes the design and communication model of the system, the design and implementation of the protocol to transmit changes in the 3D scene, the database consistency techniques and the implementation of the 3D editor interface. The current implementation status of the system shows that a real-time, full user interaction in 3D cooperative design via communication networks is achievable by our design strategy.

A real-time 3D object modeler using a small subset of implicit functions for object modeling is described in this paper. It is suitable for 3D soft object modeling. The models can be created interactively while they are being rendered. This overcomes the traditional difficulty of applying implicit functions in real-time object modeling. The modeler is implemented using object-oriented programming techniques on SGI workstations using Open Inventor. Object-oriented design patterns have been used to form the overall modeler environment. The layout of the modeler can be formed by editing it using a script language without compilation. The object models created are compatible with the VRML format.

A system that integrates 3D virtual scene generation tools and data communication techniques for 3D cooperative design work is presented. The system, named CODI (Sistema Cooperativo de Diseno3D) supports two major functions: 3D cooperative design and multiple client distant access. The system is implemented using Open Inventor Toolkit in C++ with VRML compatibility.

A novel class of generating kernels for image pyramids is introduced. When these kernels are convolved with intensity functions of images, continuous piecewise surfaces composed of polynomial tensor products are fitted to the intensity functions. The fittings are optimal in the sense that the mean square error between them and the original intensity functions is minimized. Two members of the class are introduced, and symmetry, normalization, unimodality, and equal contribution properties are proved. These kernels possess attractive properties such as small window size, fast inverse transformation, and minimum error. Experiments show that they compare favorably with existing ones in terms of mean square error.>