Ning Zhou

TL;DR: In this paper, an adaptive fractional dilated convolution (AFDC) is proposed to incorporate the information of image aspect ratios to learn more robust models, where the interpolation of nearest two integer dilated kernels are used to cope with the misalignment of fractional sampling.

TL;DR: The experimental results have demonstrated the effectiveness of the hierarchical multi-task structural learning algorithm on training more discriminative tree classifiers over the visual tree for large-scale plant species identification.

TL;DR: This algorithm can address the issues of spams, polysemes and synonyms more effectively and determine the relevance between the images and their social tags more precisely, thus it can allow us to create large amounts of training images with more reliable labels by harvesting from large-scale weakly-tagged images, which can be used to achieve more effective classifier training for many computer vision tasks.

TL;DR: An adaptive fractional dilated convolution (AFDC) is developed, which is aspect-ratio-embedded, composition-preserving and parameter-free, which can be easily implemented by common deep learning libraries and plugged into popular CNN architectures in a computation-efficient manner.