讲者简介：陈宝权，北京大学前沿计算研究中心执行主任，信息科学技术学院教授，长江学者，杰青，兼山东大学教授。纽约州立大学计算机博士。研究领域为计算机图形学与数据可视化。现任/曾任ACM TOG/IEEE TVCG编委、IEEE VIS/SIGGRAPH Asia指导委员会成员，曾任IEEE Vis 2005、ACM SIGGRAPH Asia 2014大会主席。获2003年美国NSF CAREER Award，2005年IEEE可视化国际会议最佳论文奖，和2014年中国计算机图形学杰出奖。担任973项目“城市大数据计算理论与方法”首席科学家，并任北京电影学院未来影像高精尖创新中心首席科学家。
论文题目：Numerical Coarsening using Discontinuous Basis Functions
论文作者：Jiong Chen, Hujun Bao, Tianyu Wang, Mathieu Desbrun, Jin Huang
论文摘要：In this paper, an efficient and scalable approach for simulating inhomogeneous and non-linear elastic objects is introduced. We show that numerical coarsening based on optimized non-conforming and matrix-valued shape functions allows for a more accurate simulation of heterogeneous materials with non-linear constitutive laws even on coarse grids, thus saving orders of magnitude in computational time compared to traditional nite element computations. The set of local shape functions over coarse elements is carefully tailored in a preprocessing step to balance geometric continuity and local material stiffness. In particular, we do not impose continuity of our material-aware shape functions between neighboring elements to significantly reduce the fictitious numerical stiffness that conforming bases induce; however, we require crucial geometric and physical properties such as partition of unity and exact reproduction of representative fine displacements to eschew the use of discontinuous Galerkin methods. We demonstrate that we can simulate inhomogeneous and non-linear materials significantly better than previous approaches, with no parameter tuning.
报告人邮箱：chenjiong1991 AT 126 DOT com
论文题目：Quadrangulation through Morse-Parameterization Hybridization
论文作者：Xianzhong Fang, Hujun Bao, Yiying Tong, Mathieu Desbrun, Jin Huang
论文摘要：We introduce an approach to quadrilateral meshing of arbitrary triangulated
surfaces that combines the theoretical guarantees of Morse-based approaches with the practical advantages of parameterization methods. We first construct, through an eigensolver followed by a few Gauss-Newton iterations, a periodic four-dimensional vector field that aligns with a user-provided frame field and/or a set of features over the input mesh. A field-aligned parameterization is then greedily computed along a spanning tree based on the Dirichlet energy of the optimal periodic vector field, from which quad elements are efficiently extracted over most of the surface. The few regions not yet covered by elements are then upsampled and the first component of the periodic vector field is used as a Morse function to extract the remaining quadrangles. This hybrid parameterization- and Morse-based quad meshing method is not only fast (the parameterization is greedily constructed, and the Morse function only needs to be upsampled in the few uncovered patches), but is guaranteed to provide a feature-aligned quad mesh with non-degenerate cells that closely matches the input frame field over an arbitrary surface. We show on a large variety of examples that our approach is faster than Morse-based techniques by one order of magnitude, and significantly more robust than parameterization-based techniques on models with complex features.
报告人邮箱：fxzmin AT 163 DOT com
论文作者：Ligang Liu, Chunyang Ye, Ruiqi Ni, Xiao-Ming Fu
论文摘要：We propose a novel approach, called Progressive Parameterizations, to compute foldover-free parameterizations with low isometric distortion on disk topology meshes. Instead of using the input mesh as a reference to defne the objective function, we introduce a progressive reference that contains bounded distortion to the parameterized mesh and is as close as possible to the input mesh. After optimizing the bounded distortion energy between the progressive reference and the parameterized mesh, the parameterized mesh easily approaches the progressive reference, thereby also coming close to the input. By iteratively generating the progressive reference and optimizing the bounded distortion energy to update the parameterized mesh, our algorithm achieves high-quality parameterizations with strong practical reliability and high efciency. We demonstrate that our algorithm succeeds using a massive test data set containing over 20712 complex disk topology meshes. Compared to the state-of-the-art methods, our method possesses higher computational efciency and practical reliability.
论文题目：Predictive and Generative Neural Networks for Object Functionality
论文作者：Ruizhen Hu, Zihao Yan, Jingwen Zhang, Oliver van Kaick, Ariel Shamir, Hao Zhang, Hui Huang
论文摘要：Humans can predict the functionality of an object even without any surroundings, since their knowledge and experience would allow them to “hallucinate” the interaction or usage scenarios involving the object. We develop predictive and generative deep convolutional neural networks to replicate this feat. Speciically, our work focuses on functionalities of man-made 3D objects characterized by human-object or object-object interactions. Our networks are trained on a database of scene contexts, called interaction contexts, each consisting of a central object and one or more surrounding objects, that represent object functionalities. Given a 3D object in isolation, our functional similarity network (fSIM-NET), a variation of the triplet network, is trained to predict the functionality of the object by inferring functionality-revealing interaction contexts involving the object. fSIM-NET is complemented by a generative network (iGEN-NET) and a segmentation network (iSEG-NET). iGEN-NET takes a single voxelized 3D object and synthesizes a voxelized surround, i.e., the interaction context which visually demonstrates the object’s functionalities. iSEG-NET separates the interacting objects into diferent groups according to their interaction types.
论文作者：Mingming He, Dongdong Chen, Jing Liao, Pedro V. Sander, Lu Yuan
论文摘要：We propose the first deep learning approach for exemplar-based colorization. Our network directly maps a gray scale image, together with an aligned reference color image to an output colorization. Unlike traditional exemplar-based methods to transfer color by optimizing hand-defined energies, our network learns how to select and propagate reference colors from large-scale data, which makes it robust to reference images that are similar or even irrelevant to the input image. Moreover, rather than predicting a single colorization as in other learning-based colorization methods, our network enables the user to obtain desirable results by simply feeding different references. To guide user towards efficient reference selection, the system also recommends top references with our image retrieval algorithm considering both semantic and luminance information. We validate our approach with a user study and quantitatively compare against state of the art, where we show significant improvements. Furthermore, we show our approach can be successfully extended to multi-reference and video colorization.
论文题目：Realtime Coupled Fluid/Rigid Control using Neural-Networks
论文作者：Pingchuan Ma, Yunsheng Tian, Zherong Pan, Bo Ren, Dinesh Manocha
论文摘要：We present a learning-based method to control a coupled system involving both fluid and rigid bodies. Our approach influences fluid/rigid simulator’s behavior purely at the simulation domain boundaries, leaving the rest of the domain to be governed exactly by physical laws. Compared with controllers using virtual artificial forces, our generated animations achieve higher physi- cal accuracy and visual plausibility. To solve the challenging control problem, we represent our controller using a general neural-net which is trained using deep reinforcement learning. This breaks the control task into two stages: an computationally costly training stage, and an efficient generating stage. After training, the controlled fluid animations are generated in realtime on a desktop machines by evaluating the neural net. We utilize many fluid prop- erties, e.g. the liquid’s velocity field or the smoke’s density field, to enhance the controller’s performance. We have evaluated our method on a set of complex benchmarks, where our controller drives a fluid jet to move on the domain boundary and shoot fluids towards a rigid body to accomplish a set of challenging tasks such as keeping a rigid body balanced, a two-player pingpong game, and driving a rigid body to hit a specified point on the wall.
报告人邮箱：rb AT nankai DOT edu DOT cn
报告人简介：任博于2015年于清华大学计算机科学与技术系获得工学博士学位。2015年7月至今于南开大学计算机科学与信息安全系担任讲师职位。主要研究领域与兴趣为计算机图形学中的真实感模拟、渲染方向，以及三维模型处理方向。 近年来在图形学领域国际顶级会议SIGGRAPH，顶级杂志Transactions on Graphics(TOG)等处发表文章多篇，其中在基于物理的流体模拟、渲染方向的研究获得了国际上的广泛引用与认可。目前开展的研究项目涉及三维流体模拟，真实感渲染，三维重建与几何模型处理等。
论文题目：Full 3D Reconstruction of Transparent Objects
论文作者：Bojian Wu, Yang Zhou, Yiming Qian, Minglun Gong, Hui Huang
论文摘要：Numerous techniques have been proposed for reconstructing 3D models for opaque objects in past decades. However, none of them can be directly applied to transparent objects. This paper presents a fully automatic approach for reconstructing complete 3D shapes of transparent objects. Through positioning an object on a turntable, its silhouettes and light refraction paths under different viewing directions are captured. Then, starting from an initial rough model generated from space carving, our algorithm progressively optimizes the model under three constraints: surface and refraction normal consistency, surface projection and silhouette consistency, and surface smoothness. Experimental results on both synthetic and real objects demonstrate that our method can successfully recover the complex shapes of transparent objects and faithfully reproduce their light refraction properties.
论文题目：Object-aware Guidance for Autonomous Scene Reconstruction
论文作者：Ligang Liu, Xi Xia, Han Sun, Qi Shen, Juzhan Xu, Bin Chen, Hui Huang, Kai Xu
论文摘要：Autonomous 3D scene scanning and reconstruction of unknown indoor scenes by mobile robots with depth sensors has become an active research area in recent years. However, it suffers the problem of balancing between global exploration of the scene and local scanning of the objects. In this paper, we propose an object-aware guidance autoscanning approach for on-the-fly exploration, reconstruction, and understanding of unknown scenes in one navigation pass. Our approach interleaves between object analysis for identifying next best object (NBO) for global exploration, and object-aware information gain analysis for planning next best view (NBV) for local scanning. Based on a model-driven objectness measurement, an objectness based segmentation method is introduced to extract semantic object proposals in the current scene surface via a multi-class graph cuts minimization. Then we propose objectness based NBO and NBV strategies to plan both global navigation path and local scanning views. An object of interest (BOI) is identified by the NBO metric determined by both its objectness score and visual saliency. The robot then moves and visit the BOI and conducts the scanning with views provided by the NBV strategy. When the BOI is recognized as a complete object, the most similar 3D model in the dataset is inserted into the scene to replace it. The algorithm iterates until all objects are recognized and reconstructed in the scene. A variety of experiments and comparisons have shown the feasibility and efficiency of our proposed approach.
论文题目：Creating and Chaining Camera Moves for Quadrotor Videography
论文作者：Ke Xie, Hao Yang, Shengqiu Huang, Dani Lischinski, Marc Christie, Kai Xu, Minglun Gong, Daniel Cohen-Or, Hui Huang
论文摘要：Capturing aerial videos with a quadrotor-mounted camera is a challenging creative task, as it requires the simultaneous control of the quadrotor’s position and the mounted camera’s orientation. Letting the drone follow a pre-planned trajectory is a much more appealing option, and recent research has proposed a number of tools designed to automate the generation of feasible camera motion plans; however, these tools require the user to specify and edit the camera path, for example by providing a complete and ordered sequence of key viewpoints.
In this paper, we propose a higher level tool designed to enable even novice users to easily capture compelling aerial videos of large scale outdoor scenes. Using a coarse 2.5D model of a scene, the user is only expected to specify starting and ending viewpoints and designate a set of landmarks, with or without a particular order. Our system automatically generates a diverse set of candidate local camera moves for observing each landmark, which are collision-free, smooth, and adapted to the shape of the landmark. These moves are guided by a landmark-centric view quality field, which combines visual interest and frame composition. An optimal global camera trajectory is then constructed that chains together a sequence of local camera moves, by choosing one move for each landmark and connecting them with suitable transition trajectories. This task is formulated and solved as an instance of the Set Traveling Salesman Problem.
论文摘要：We present an automatic algorithm for subtractive manufacturing of freeform 3D objects using high-speed CNC machining. A CNC machine operates a cylindrical drill to carve off material from a 3D shape stock, following a tool path, to ``expose'' the target object. Our method decomposes the input object's surface into a small number of patches each of which is fully accessible and machinable by the CNC machine, in continuous fashion, under a fixed drill-object setup configuration. This is achieved by covering the input surface using a minimum number of accessible regions and then extracting a set of machinable patches from each accessible region. For each patch obtained, we compute a continuous, space-filling, and iso-scallop tool path which conforms to the patch boundary, enabling efficient carving with high-quality surface finishing. The tool path is generated in the form of connected Fermat spirals, which have been generalized from a 2D fill pattern for layered manufacturing to work for curved surfaces. Furthermore, we develop a novel method to control the spacing of Fermat spirals based on directional surface curvature and adapt the heat method to obtain iso-scallop carving. We demonstrate automatic generation of accessible and machinable surface decompositions and iso-scallop Fermat spiral carving paths for freeform 3D objects. Comparisons are made to commercially available tool paths in terms of real CNC machining time and surface quality.
报告人邮箱：haisenzhao AT gmail DOT com
报告人简介：山东大学计交叉研究中心博士生，师从陈宝权教授。主要研究方向为智能制造相关的计算机图形学，多篇文章发表于SIGGRAPH和 Transaction on Graphics上，在软件学报和 Pacific Vis发表论文各一篇。曾荣获“山东大学校长奖学金”，“CAD&CG 2012优秀学生论文”等荣誉。
论文摘要：Many computer graphics problems require computing geometric shapes subject to certain constraints. This often results in non-linear and non-convex optimization problems with globally coupled variables, which pose great challenge for interactive applications. Local-global solvers developed in recent years can quickly compute an approximate solution to such problems, making them an attractive choice for applications that prioritize efficiency over accuracy. However, these solvers suffer from lower convergence rate, and may take a long time to compute an accurate result. In this paper, we propose a simple and effective technique to accelerate the convergence of such solvers. By treating each local-global step as a fixed-point iteration, we apply Anderson acceleration, a well-established technique for fixed-point solvers, to speed up the convergence of a local-global solver. To address the stability issue of classical Anderson acceleration, we propose a simple strategy to guarantee the decrease of target energy and ensure its global convergence. In addition, we analyze the connection between Anderson acceleration and quasi-Newton methods, and show that the canonical choice of its mixing parameter is suitable for accelerating local-global solvers. Moreover, our technique is effective beyond classical local-global solvers, and can be applied to iterative methods with a common structure. We evaluate the performance of our technique on a variety of geometry optimization and physics simulation problems. Our approach significantly reduces the number of iterations required to compute an accurate result, with only a slight increase of computational cost per iteration. Its simplicity and effectiveness makes it a promising tool for accelerating existing algorithms as well as designing efficient new algorithms.
论文摘要：The real world exhibits an abundance of non-stationary textures. Examples include textures with large scale structures, as well as spatially variant and inhomogeneous textures. While existing example-based texture synthesis methods can cope well with stationary textures, non-stationary textures still pose a considerable challenge, which remains unresolved. In this paper, we propose a new approach for example-based non-stationary texture synthesis. Our approach uses a generative adversarial network (GAN), trained to double the spatial extent of texture blocks extracted from a specifc texture exemplar. Once trained, the fully convolutional generator is able to expand the size of the entire exemplar, as well as of any of its sub-blocks. We demonstrate that this conceptually simple approach is highly eﬀective for capturing large scale structures, as well as other non-stationary attributes of the input exemplar. As a result, it can cope with challenging textures, which, to our knowledge, no other existing method can handle.
论文摘要：Correspondence between images is a fundamental problem in computer vision, with a variety of graphical applications. In this talk we will present a new method that demonstrates the abilities of deep features of classification network to precisely localize corresponding points between two images.
Our method is designed for pairs of images where the main objects of interest may belong to different semantic categories and differ drastically in shape and appearance, yet still contain semantically related or geometrically similar parts.The usefulness of our method is demonstrated using a variety of graphics applications, including cross-domain image alignment, creation of hybrid images, automatic image morphing, and more.
报告人：Kfir Aberman, 北京电影学院未来影像高精尖创新中心
报告人邮箱：kfirab DOT t2 AT gmail DOT com
报告人简介：Kfir is an Israeli researcher in the advanced innovation center for future visual entertainment (AICFVE) located at the Beijing Film Academy. His areas of interests include deep neural network architectures and their applications in computer graphics. Kfir has experience of several years in computer vision, analyzation of visual data and visual effects as an algorithm team leader in the Israeli defense Intelligence (IDI). Kfir holds a B.Sc. (summa cum laude) and M.Sc. (cum laude) in electrical engineering from the Technion and is pursuing his PhD in Tel-Aviv University.