Shawn Newsam

TL;DR: This work considers a standard non-spatial representation in which the frequencies but not the locations of quantized image features are used to discriminate between classes analogous to how words are used for text document classification without regard to their order of occurrence, and considers two spatial extensions.

TL;DR: An extensive evaluation of local invariant features for image retrieval of land-use/land-cover classes in high-resolution aerial imagery using a bag-of-visual-words (BOVW) representation and describes interesting findings such as the performance-efficiency tradeoffs that are possible through the appropriate pairings of different-sized codebooks and dissimilarity measures.

TL;DR: It is found that FG domains are structurally and chemically heterogeneous, and adopt distinct categories of intrinsically disordered structures in non-random distributions.

TL;DR: A novel image representation termed spatial pyramid co-occurrence which characterizes both the photometric and geometric aspects of an image which achieves state-of-the-art performance on the Graz-01 object class dataset and performs competitively on the 15 Scene dataset.

TL;DR: In this article, a video prediction-based methodology was proposed to scale up training sets by synthesizing new training samples in order to improve the accuracy of semantic segmentation networks, which achieved state-of-the-art mIoUs of 83.5% on Cityscapes and 82.9% on CamVid.