Search

As re-ranking is a necessary procedure to boost person re-identification (re-ID) performance on large-scale datasets, the diversity of feature becomes crucial to person reID for its importance both on designing pedestrian descriptions and re-ranking based on feature fusion. However, in many circumstances, only one type of pedestrian feature is available. In this paper, we propose a "Divide and use" re-ranking framework for person re-ID. It exploits the diversity from different parts of a high-dimensional feature vector for fusion-based re-ranking, while no other features are accessible. Specifically, given an image, the extracted feature is divided into sub-features. Then the contextual information of each sub-feature is iteratively encoded into a new feature. Finally, the new features from the same image are fused into one vector for re-ranking. Experimental results on two person re-ID benchmarks demonstrate the effectiveness of the proposed framework. Especially, our method outperforms the state-of-the-art on the Market-1501 dataset., Comment: Accepted by BMVC2017

We investigate the SU($2N$) symmetry effects with $2N>2$ on the two-dimensional interacting Dirac fermions at finite temperatures, including the valence-bond-solid transition, the Pomeranchuk effect, the compressibility and the uniform spin susceptibility, by performing the determinant quantum Monte Carlo simulations of the half-filled SU($2N$) Hubbard model on a honeycomb lattice. The columnar valence-bond-solid (cVBS) phase only breaks the three-fold discrete symmetry, and thus can survive at finite temperatures. The disordered phase in the weak coupling regime is the thermal Dirac semi-metal state, while in the strong coupling regime it is largely a Mott state in which the cVBS order is thermally melted. The calculated entropy-temperature relations for various values of the Hubbard interaction $U$ show that, the Pomeranchuk effect occurs when the specific entropy is below a characteristic value of $S^*$ --- the maximal entropy per particle from the spin channel of local moments. The SU($2N$) symmetry enhances the Pomeranchuk effect, which facilitates the interaction-induced adiabatic cooling. Our work sheds new light on future explorations of novel states of matter with ultra-cold large-spin alkaline fermions., Comment: 12 pages, 15 figures

Projective analysis is an important solution for 3D shape retrieval, since human visual perceptions of 3D shapes rely on various 2D observations from different view points. Although multiple informative and discriminative views are utilized, most projection-based retrieval systems suffer from heavy computational cost, thus cannot satisfy the basic requirement of scalability for search engines. In this paper, we present a real-time 3D shape search engine based on the projective images of 3D shapes. The real-time property of our search engine results from the following aspects: (1) efficient projection and view feature extraction using GPU acceleration; (2) the first inverted file, referred as F-IF, is utilized to speed up the procedure of multi-view matching; (3) the second inverted file (S-IF), which captures a local distribution of 3D shapes in the feature manifold, is adopted for efficient context-based re-ranking. As a result, for each query the retrieval task can be finished within one second despite the necessary cost of IO overhead. We name the proposed 3D shape search engine, which combines GPU acceleration and Inverted File Twice, as GIFT. Besides its high efficiency, GIFT also outperforms the state-of-the-art methods significantly in retrieval accuracy on various shape benchmarks and competitions., Comment: accepted by CVPR16, achieved the first place in Shrec2016 competition: Large-Scale 3D Shape Retrieval under the perturbed case

The interplay between charge and spin degrees of freedom in strongly correlated fermionic systems, in particular of Dirac fermions, is a long-standing problem in condensed matter physics. We investigate the competing orders in the half-filled SU(2N) Hubbard model on a honeycomb lattice, which can be accurately realized in optical lattices with large-spin ultra-cold alkaline-earth fermions. Employing large-scale projector determinant quantum Monte Carlo simulations, we have explored quantum phase transitions from the gapless Dirac semi-metals to the gapped Mott-insulating phases in the SU(4) and SU(6) cases. Both of these Mott-insulating states are found to be columnar valence bond solid (cVBS) and to be absent of the antiferromagnetic Neel ordering and the loop current ordering. Inside the cVBS phases, the dimer ordering is enhanced by increasing fermion components and behaves non-monotonically as the interaction strength increases. Although the transitions generally should be of first order due to a cubic invariance possessed by the cVBS order, the coupling to gapless Dirac fermions can soften the transitions to second order through a non-analytic term in the free energy. Our simulations provide important guidance for the experimental exploration of novel states of matter with ultra-cold alkaline earth fermions., Comment: 20 pages, 26 figures

We have systematically studied the thermodynamic properties of a two-dimensional half-filled SU(2N) Hubbard model on a square lattice by using the determinant quantum Monte Carlo method. The entropy-temperature relation, the isoentropy curve, and the probability distribution of the onsite occupation number are calculated in both SU(4) and SU(6) cases, which exhibit prominent features of the Pomeranchuk effect. We analyze these thermodynamic behaviors based on charge and spin energy scales. In the charge channel, the interaction strength that marks the crossover from the weak to strong interaction regimes increases with the number of fermion components. In the spin channel, increasing the number of fermion components enhances quantum spin fluctuations, which is shown in the simulations of uniform spin susceptibilities and antiferromagnetic structure factors., Comment: 10 pages, 6 figures