This paper considers vehicle re-identification (re-ID) problem. The extreme viewpoint variation (up to 180 degrees) poses great challenges for existing approaches. Inspired by the behavior in human's recognition process, we propose a novel viewpoint-aware metric learning approach. It learns two metrics for similar viewpoints and different viewpoints in two feature spaces, respectively, giving rise to viewpoint-aware network (VANet). During training, two types of constraints are applied jointly. During inference, viewpoint is firstly estimated and the corresponding metric is used. Experimental results confirm that VANet significantly improves re-ID accuracy, especially when the pair is observed from different viewpoints. Our method establishes the new state-of-the-art on two benchmarks. 12 * Corresponding author 1 R. Chu and Y. Sun share equal contribution. 2 Work done at Megvii Technology.

We map United States comparative advantage between 1980 and 1995, by trading partner and region, using Balassa's export-based index of Revealed Comparative Advantage (RCA). We find: temporally stable and ubiquitous US comparative advantage in differentiated producer goods (except disadvantage in Japan); somewhat less stable and less sweeping US disadvantage in standardized producer goods; chaotic and diverse patterns of US RCA in consumer goods (especially in the Chinese market). Our most significant findings are surprisingly sharp geographical differences in patterns of US RCA and surprisingly small differences across sub-sectors of 1, 2, and 3-digit SITC classificationsregional, but not sectoral, "niche" specialization. The high overall variability across regions in RCA indexes seems unrelated to obvious explanations such as proximity or lingual/historical ties to the US. In producer goods, RCA variability across regions correlates somewhat better with accounts of trade diversion and of regional preferences for and discrimination against US exports. We find only scant evidence of high or increasing variability across disaggregated commodity subgroups in US RCA indexes. Such variability is often the prediction of theories of comparative advantage that are based on vertical specialization, product differentiation, or scale and agglomeration economies.

No abstract

No abstract

Imposing smoothness priors is a key idea of the top-ranked global stereo models. Recent progresses demonstrated the power of second order priors which are usually defined by either explicitly considering three-pixel neighborhoods, or implicitly using a so-called 3D-label for each pixel. In contrast to the traditional first-order priors which only prefer fronto-parallel surfaces, second-order priors encourage arbitrary collinear structures. However, we still can find defective regions in matching results even under such powerful priors, e.g., large textureless regions. One reason is that most of the stereo models are non-convex, where pixelwise smoothness priors, i.e., local constraints, are too flexible to prevent the solution from trapping in bad local minimums. On the other hand, long-range spatial constraints, especially the segment-based priors, have advantages on this problem. However, segment-based priors are too rigid to handle curved surfaces. We present a mixture model to combine the benefits of these two kinds of priors, whose energy function consists of two terms 1) a Laplacian operator on the disparity map which imposes pixel-wise second-order smoothness; 2) a segment-wise matching cost as a function of quadratic surface, which encourages "as-rigid-as-possible" smoothness. To effectively solve the problem, we introduce an intermediate term to decouple the two subenergies, which enables an alternated optimization algorithm that is about an order of magnitude faster than PatchMatch [1]. Our approach is one of the top ranked models on the Middlebury benchmark at sub-pixel accuracy.