Jun Yu

TL;DR: This paper consists of the following contributions: massive social images and their privacy settings are leveraged to learn the object-privacy relatedness effectively and identify a set of privacy-sensitive object classes automatically and a deep multi-task learning algorithm is developed.

TL;DR: The proposed model defeats the state-of-the-art deep learning approaches applied to place recognition and is easily trained via the standard backpropagation method.

TL;DR: The proposed HD-MSL effectively combines varied features into a unified representation and integrates the labeling information based on a probabilistic framework and can automatically learn a combination coefficient for each view, which plays an important role in utilizing the complementary information of multiview data.

TL;DR: This approach improves traditional methods by adopting multiview locality-sensitive sparse coding in the retrieving process, and incorporates a local similarity preserving term into the objective of sparse coding, which groups similar silhouettes to alleviate the instability of sparse codes.

TL;DR: A semisupervised multiview distance metric learning (SSM-DML) that can simultaneously accomplish cartoon character classification and dissimilarity measurement is proposed and developed.