No abstract

Figure 1: Blind visual motif removal results on images unseen during training. Top: test images embedded with semitransparent motifs. Bottom: our reconstructed results. Our network was trained on Latin characters, yet successfully identifies and removes the Hindi and Japanese characters (left three images). Similarly, the overlaid visual motifs on the right three images differ semantically from the motifs used during training. AbstractMany images shared over the web include overlaid objects, or visual motifs, such as text, symbols or drawings, which add a description or decoration to the image. For example, decorative text that specifies where the image was taken, repeatedly appears across a variety of different images. Often, the reoccurring visual motif, is semantically similar, yet, differs in location, style and content (e.g., text placement, font and letters). This work proposes a deep learning based technique for blind removal of such objects. In the blind setting, the location and exact geometry of the motif are unknown. Our approach simultaneously estimates which pixels contain the visual motif, and synthesizes the underlying latent image. It is applied to a single input image, without any user assistance in specifying the location of the motif, achieving state-of-the-art results for blind removal of both opaque and semi-transparent visual motifs.We present a method for completely blind visual motif removal. In the blind setting, the exact location, structure and size of these motifs is unknown. The generalization ability of our network is demonstrated by removing visual motifs that are not seen during training, e.g., watermark re-

The accurate representation of two-dimensional images in three dimensions has become important for many medical imaging applications and for cardiac magnetic resonance imaging (MRI) in particular. Reconstruction methods applied after data acquisition can produce three-dimensional information from two-dimensional data and make applications such as surgical planning more effective. Current reconstruction techniques usually demand contrast agents, and can suffer due to poor segmentation and sampling constraints that cause surface irregularities and distort dimensions. The novel technique presented here for anatomical modeling uses adaptive control grid interpolation (ACGI) to approximate data not captured by scanning, and a progressive shape-element segmentation technique to complete reconstruction. Quantitative validations conducted on models of pediatric cardiac malformations have confirmed the theoretical advantages of this technique, and that higher quality is achieved than with competing methods based on geometric parameters. Vascular diameters from reconstructions showed errors of less than 1% for a known geometry as compared to over 9% for competing methods. Qualitatively, models produced with the new methodology displayed substantial improvement over alternatives. Approximately 50 rare cardiac structures, including surgically altered Fontan and atypical aortic anatomies, have been reconstructed. All data used to create these reconstructions were acquired using standard pulse sequences and without contrast agents. Benefits of the new technique are particularly evident when complex vascular configurations complicate reconstruction. The proposed methodology enables a powerful tool allowing physicians to analyze and manipulate highly accurate and clearly presented vascular structures in an interactive medium.

No abstract

Fig. 1. Employing deep embeddings for clustering 3D shapes. Above, we use PCA to visualize the output embedding of point clouds of chairs. We also highlight (in unique colors) a few random clusters and display a few representative chairs from these clusters.Recently, there has been increasing interest to leverage the competence of neural networks to analyze data. In particular, new clustering methods that employ deep embeddings have been presented. In this paper, we depart from centroid-based models and suggest a new framework, called Clustering-driven deep embedding with PAirwise Constraints (CPAC), for non-parametric clustering using a neural network. We present a clusteringdriven embedding based on a Siamese network that encourages pairs of data points to output similar representations in the latent space. Our pair-based model allows augmenting the information with labeled pairs to constitute a semi-supervised framework. Our approach is based on analyzing the losses associated with each pair to refine the set of constraints. We show that clustering performance increases when using this scheme, even with a limited amount of user queries. We demonstrate how our architecture is adapted for various types of data and present the first deep framework to cluster 3D shapes.