Search

Deformation transfer is a type of retargeting method that operates directly on the mesh and, by doing so, enables reuse of animation without setting up character rigs and a mapping between the source and target geometries. Deformation transfer can potentially reduce the costs of animation and give studios a competitive edge when keeping up with the latest computer animation technology. Unfortunately, deformation transfer has limitations and is yet to become standard practice in the industry. This survey starts by introducing Sumner and Popovic’s [18] seminal work and highlights key issues for industry settings. We then review related work in sections, organized by these key issues. After surveying related work, we discuss how their advances open the door to several practical applications of deformation transfer. To conclude, we highlight areas of future work.

We propose an interactive video completion method aiming for practical use in a digital production workplace. The results of earlier automatic solutions often require a considerable amount of manual modifications to make them usable in practice. To reduce such a laborious task, our method offers an efficient editing tool. Our iterative algorithm estimates the flow fields and colors in space-time holes in the video. As in earlier approaches, our algorithm uses an $L^1$L1 data term to estimate flow fields. However, we employ a novel $L^2$L2 data term to estimate temporally coherent color transitions. Our graphics processing unit implementation enables the user to interactively complete a video by drawing holes and immediately removes objects from the video. In addition, our method successfully interpolates sparse modifications initialized by the designer. According to our subjective evaluation, the videos completed with our method look significantly better than those with other state-of-the-art approaches.

We present an algorithm for efficient rendering of animated hair under a dynamic, low-frequency lighting environment. We use spherical harmonics (SH) to represent the environmental light. The transmittances between a point on a hair strand and the light sources are also represented by SH functions. Then, a convolution of SH functions and the scattering function of a hair strand is precomputed. This allows us to efficiently compute the intensity at a point on the hair. However, the computation of the transmittance is very time-consuming. We address this problem by using a voxel-based approach: the transmittance is computed by using a voxelized hair model. We further accelerate the computation by sampling the voxels. By using our method, we can render a hair model consisting of tens of thousands of hair strands at interactive frame rates.

Recent works in image editing are opening up new possibilities to manipulate and enhance input images. Within this context, we leverage well-known characteristics of human perception along with a simple depth approximation algorithm to generate non-photorealistic renditions that would be difficult to achieve with existing methods. Once a perceptually plausible depth map is obtained from the input image, we show how simple algorithms yield powerful new depictions of such an image. Additionally, we show how artistic manipulation of depth maps can be used to create novel non-photorealistic versions, for which we provide the user with an intuitive interface. Our real-time implementation on graphics hardware allows the user to efficiently explore artistic possibilities for each image. We show results produced with six different styles proving the versatility of our approach, and validate our assumptions and simplifications by means of a user study.

In a cel animation workplace, non-photorealistic rendering (NPR) techniques are often applied to a 3D scene model, which is constructed by referring to the 2D hand-drawn images of the scene. There exist a lot of scene features that must be modeled from the 2D information. The accentuation effects are amongst the most importantfeatures in a hand-drawn scene. In this context, accentuation refers to adding hand-drawn shading of the scene. Thespecific region that an artist wishes the viewer to observe is drawn more brightly than the other regions. We propose a method of extracting the accentuation effects from a single 2D painted image of an animated scene. This method enables the automatic addition of the accentuation effects to each frame of the animation, observed from other viewpoints.

The multivariate normal is widely used as the expected distribution of face shape. It has been used for face detection and tracking in computer vision, as a prior for facial animation editing in computer graphics, and as a model in psychological theory. In this contribution we consider the character of the multivariate normal in high dimensions, and show that these applications are not justified. While we provide limited evidence that facial proportions are not Gaussian, this is tangential to our conclusion: even if faces are truly “Gaussian”, maximum a posteriori and other applications and conclusions that assume that typical faces lie near the mean are not valid.

We report a method for reconstructing a three-dimensional (3D) depth map using a single two-dimensional (2D) image. Our method is designed to reconstruct manmade objects, such as buildings. We first estimate the normal map, and then integrate it to obtain the depth map. To estimate the normal map, we analyze the co-occurrence relation between the normal vectors and image features in a training dataset. We consider the corners and lines to be the image features. The training dataset is formed of 3D game scenes. In the offline learning phase, we detect corners and lines in the normal map of each game scene using a detection algorithm, and observe the normal vectors around them. Then we construct a database of the co-occurrence relations, i.e., how frequently each corner or line appears with each normal vector. In the online reconstruction phase, given an input image, we detect the corners and lines using the same detection algorithm, and estimate the normal vectors around them based on the learned co-occurrence relation. We formulate this estimation using a Markov random field. Finally, the estimated normal map is integrated by solving Poisson’s equation, and we obtain a depth map.

International audience; We present a novel approach to the creation of varied animations from a small set of simple 2D input shapes. Instead of providing a new 2D shape for each keyframe of an animation sequence, we instead interpolate between a few example shapes in a reduced pose-space. Similar approaches have been presented in the past, but were restricted in the types of input or range of deformations allowed. In order to address these limitations, we reformulate the problem as an N-way morphing process on 2D input bitmap or vector graphics. Our formulation includes an N-way mapping technique, an efficient, rigidity preserving non-linear blending function, improved extrapolation, and a novel scattered data interpolation technique to manage the reduced pose-space. The resulting animations are correlated to paths in the reduced pose-space, allowing users to intuitively and interactively control temporal behaviors with simple gestures. We demonstrate our techniques in several example animations.