Saurabh Singh — Page 7

arXiv2013

Ancient eclipses and long-term drifts in the Earth - Moon system

We investigate the anomalies in the Earth - Moon system using ancient eclipse data. We identify nine groups of anomalous eclipses between 400 and 1800 AD recorded in parts of India that should have completely missed the subcontinent as per NASA simulations (Espenak and Meeus, 2011). We show that the typical correction to the lunar location required to reconcile the anomalous eclipses is relatively small and consistent with the fluctuations in the length of day that are observed in recent periods. We then investigate the change in Earth's moment of inertia due to differential acceleration of land and water that can account for this discrepancy. We show that 80 percent of these discrepancies occur when the Moon is at declinations greater than 10 deg and closer to its major standstill of 28 deg while it spends 46 percent of the time in this region. We simulate the differential interaction of the Moon's gravity with landmass and water using finite element method to account for landmass and water mass. We show that the results of eclipse error are consistent with the estimate of a small differential acceleration when the Moon is over land at high latitudes. However, we encounter some examples where the results from simulations studies cannot explain the phenomena. Hence we propose that dT corrections have to be coupled with some other mechanism possibly a small vertical oscillation in the Moon's rotational plane with a period of the order of a few hundred years to achieve the required adjustment in the Eclipse maps.

astro-ph.EP↗

arXiv2012

Unsupervised Discovery of Mid-Level Discriminative Patches

The goal of this paper is to discover a set of discriminative patches which can serve as a fully unsupervised mid-level visual representation. The desired patches need to satisfy two requirements: 1) to be representative, they need to occur frequently enough in the visual world; 2) to be discriminative, they need to be different enough from the rest of the visual world. The patches could correspond to parts, objects, "visual phrases", etc. but are not restricted to be any one of them. We pose this as an unsupervised discriminative clustering problem on a huge dataset of image patches. We use an iterative procedure which alternates between clustering and training discriminative classifiers, while applying careful cross-validation at each step to prevent overfitting. The paper experimentally demonstrates the effectiveness of discriminative patches as an unsupervised mid-level visual representation, suggesting that it could be used in place of visual words for many tasks. Furthermore, discriminative patches can also be used in a supervised regime, such as scene classification, where they demonstrate state-of-the-art performance on the MIT Indoor-67 dataset.

cs.CV↗