• Citations Per Year
Learn More
BACKGROUND The upper lip can be divided into aesthetic subunits. Accordingly, we classified the upper lip defect into 3 categories: medial subunit defect, lateral subunit defect, and cross-subunit defect. The defect should be repaired aesthetically. We developed an innovative partial-thickness myocutaneous flap based on the vascular network of the(More)
PURPOSE This study aims to investigate the 3-dimensional (3D) anatomical structure of the orbicularis oris and nasalis, which are closely associated with the appearance of the upper lip and lower part of the nose. The relationship of the complicated 3D anatomical structure with the outline shape was also determined. METHODS Microcomputed tomography(More)
A thorough knowledge of the anatomic structure of the orbicularis oris of the upper lip and the nasalis in fetus with cleft lip is the key for the success of cleft lip repair. To understand the anatomic structure of the muscles of nasolabial region in fetus with cleft lip, the nasolabial tissues in 4 aborted fetuses with cleft lip were soaked for 7 days(More)
BACKGROUND In microform cleft lip repair, reconstructing the elaborate structures is difficult. We describe a new technique of unilateral microform cleft lip repair that is based on the muscle tension line group theory. METHODS According to the shape of Cupid bow, a different small incision is used without creating an obvious cutaneous scar. First, the(More)
BACKGROUND Plastic surgeons have attempted various ways to rebuild the aesthetic subunits of the upper lip in patients with cleft lip with less than perfect results in most cases. We propose that repairing the 3 muscle tension line groups in the nasolabial complex will have improved aesthetic results. METHODS Micro-computed tomographic scans were(More)
We study the problem of learning generative models of 3D shapes. Voxels or 3D parts have been widely used as the underlying representations to build complex 3D shapes; however, voxel-based representations suffer from high memory requirements, and parts-based models require a large collection of cached or richly parametrized parts. We take an alternative(More)
Figures 1, 2, 3, and 4 show more experimental results on randomly sampled floorplan images in our dataset, demonstrating their challenges and varying styles. Figures 5 and 6 show our results on other image sources, in particular, Rent3D dataset [2] and Google Image Search, respectively. Fig. 7 shows some typical failure cases. Lastly, the main paper has(More)
Inference of correspondences between images from different modalities is an extremely important perceptual ability that enables humans to understand and recognize crossmodal concepts. In this paper, we consider an instance of this problem that involves matching photographs of building interiors with their corresponding floorplan. This is a particularly(More)