This paper presents a data-driven approach to simultaneous segmentation and labeling of parts in 3D meshes. An objective function is formulated as a Conditional Random Field model, with terms assessing the consistency of faces with labels, and terms between labels of neighboring faces. The objective function is learned from a collection of labeled training meshes. The algorithm uses hundreds of geometric and contextual label features and learns different types of segmentations for different tasks, without requiring manual parameter tuning. Our algorithm achieves a significant improvement in results over the state-of-the-art when evaluated on the Princeton Segmentation Benchmark, often producing segmentations and labelings comparable to those produced by humans.

No abstract

We present a method for browsing videos by directly dragging their content. This method brings the benefits of direct manipulation to an activity typically mediated by widgets. We support this new type of interactivity by: 1) automatically extracting motion data from videos; and 2) a new technique called relative flow dragging that lets users control video playback by moving objects of interest along their visual trajectory. We show that this method can outperform the traditional seeker bar in video browsing tasks that focus on visual content rather than time.

No abstract

Figure 1: Our analytic drawing tool infers 3D scaffolds of linear segments (a) from sketched strokes. 3D feature curves can then be sketched by deriving position and tangent constraints from the scaffold (b). After fixing the viewpoint and adding image-space silhouette curves (c), we apply traditional hand-rendering techniques [Robertson 2003] to create a production design drawing of the espresso machine (d). AbstractWe describe a novel approach to inferring 3D curves from perspective drawings in an interactive design tool. Our methods are based on a traditional design drawing style known as analytic drawing, which supports precise image-space construction of a linear 3D scaffold. This scaffold in turn acts as a set of visual constraints for sketching 3D curves. We implement analytic drawing techniques in a pure-inference sketching interface which supports both singleand multi-view incremental construction of complex scaffolds and curve networks. A new representation of 3D drawings is proposed, and useful interactive drawing aids are described. Novel techniques are presented for deriving constraints from single-view sketches drawn relative to the current 3D scaffold, and then inferring 3D line and curve geometry which satisfies these constraints. The resulting analytic drawing tool allows 3D drawings to be constructed using exactly the same strokes as one would make on paper.