Interactive multigraph visualization and exploration with a two-phase strategy

While it is very reasonable to use a multigraph consisting of multiple edges between vertices to represent various relationships, the multigraph has not drawn much attention in research. To visualize such a multigraph, a clear layout representing a global structure is of great importance, and interactive visual analysis which allows the multiple edges to be adjusted in appropriate ways for detailed presentation is also essential. A novel interactive two-phase approach to visualizing and exploring multigraph is proposed. The approach consists of two phases： the first phase improves the previous popular works on force-directed methods to produce a brief drawing for the aggregation graph of the input multigraph, while the second phase proposes two interactive strategies, the magnifier model and the thematic-oriented subgraph model. The former highlights the internal details of an aggregation edge which is selected interactively by user, and draws the details in a magnifying view by cubic Bezier curves; the latter highlights only the thematic subgraph consisting of the selected multiple edges that the user concerns. The efficiency of the proposed approach is demonstrated with a real-world multigraph dataset and how it is used effectively is discussed for various potential applications.

A novel unsupervised approach for multilevel image clustering from unordered image collection

A novel unsupervised approach to automatically constructing multilevel image clusters from unordered im- ages is proposed in this paper. The whole input image col- lection is represented as an imaging sample space （ISS） con- sisting of globally indexed image features extracted by a new efficient multi^view image feature matching method. By mak- ing an analogy between image capturing and observation of ISS, each image is represented as a binary sequence, in which each bit indicates the visibility of a corresponding feature. Based on information theory-inspired image popularity and dissimilarity measures, we show that the image content and distance can be quantitatively described, guided by which an input image collection is organized into multilevel clusters automatically. The effectiveness and the efficiency of the pro- posed approach are demonstrated using three real image col- lections and promising results were obtained from both qual- itative and quantitative evaluation.

Depth estimation using an improved stereo network

Self-supervised depth estimation approaches present excellent results that are comparable to those of the fully supervised approaches,by employing view synthesis between the target and reference images in the training data.ResNet,which serves as a backbone network,has some structural deficiencies when applied to downstream fields,because its original purpose was to cope with classification problems.The low-texture area also deteriorates the performance.To address these problems,we propose a set of improvements that lead to superior predictions.First,we boost the information flow in the network and improve the ability to learn spatial structures by improving the network structures.Second,we use a binary mask to remove the pixels in low-texture areas between the target and reference images to more accurately reconstruct the image.Finally,we input the target and reference images randomly to expand the dataset and pre-train it on ImageNet,so that the model obtains a favorable general feature representation.We demonstrate state-of-the-art performance on an Eigen split of the KITTI driving dataset using stereo pairs.