Neotropical refugia

Patterns of endemism in the Neotropics have been explained by restriction of forest to ‘refugia’ in arid cold-stages of the Quaternary (Haffer J (1969)
Speciation in Amazonian forest birds. Science 165: 131–137). The palaeoecological record, however, shows no such forest contraction. We review
palaeoecological and phylogenetic data on the response of Neotropical taxa and communities to climatic changes of the Cenozoic. Solar insolation varies
over this period with latitude and geography, including shifts in opposite directions between high and low latitudes. In the Neotropics, distribution and
abundance patterns originate on a wide range of timescales through the Cenozoic, down to the currently dominant precession forcing (20 kyr). In contrast,
distributions and abundances at higher latitudes are controlled by obliquity forcing (40 kyr). The patterns observed by Haffer (1969) are likely derived
from pre-Quaternary radiations and are not inconsistent with palaeoecological findings of continuous forest cover in major areas of the Neotropics
during the Quaternary. The relative proportions of speciation processes have changed through time between predominantly sympatric to predominantly
allopatric depending on the prevailing characteristics of orbitally forced climatic changes. Behaviour of Neotropical organisms and ecosystems on long
timescales may be influenced much more by precessional forcing than by the obliquity forcing that controls high-latitude climate change and glaciations.

A spatio-temporal 2D-models framework for human pose recovery in monocular sequences

This paper addresses the pose recovery problem of a particular articulated object: the human body. In this model-based approach, the 2D-shape is associated to the corresponding stick figure allowing the joint segmentation and pose recovery of the subject observed in the scene. The main disadvantage of 2D-models is their restriction to the viewpoint. To cope with this limitation, local spatio-temporal 2D-models corresponding to many views of the same sequences are trained, concatenated and sorted in a global framework. Temporal and spatial constraints are then considered to build the probabilistic transition matrix (PTM) that gives a frame to frame estimation of the most probable local models to use during the fitting procedure, thus limiting the feature space. This approach takes advantage of 3D information avoiding the use of a complex 3D human model. The experiments carried out on both indoor and outdoor sequences have demonstrated the ability of this approach to adequately segment pedestrians and estimate their poses independently of the direction of motion during the sequence. (c) 2008 Elsevier Ltd. All rights reserved.