A5056489323- Physics of Superconductivity and Magnetism
- Quantum many-body systems
- Quantum and electron transport phenomena
- Face and Expression Recognition
- Face recognition and analysis
- Theoretical and Computational Physics
- Quantum Chromodynamics and Particle Interactions
- Date Palm Research Studies
- Video Coding and Compression Technologies
- Forest Insect Ecology and Management
- Cold Atom Physics and Bose-Einstein Condensates
- Advanced Condensed Matter Physics
- Advanced Vision and Imaging
- Magnetic and transport properties of perovskites and related materials
- Insect and Arachnid Ecology and Behavior
- Indoor and Outdoor Localization Technologies
- Computer Graphics and Visualization Techniques
Beijing Forestry University
2024
East China Normal University
2018
Rutgers, The State University of New Jersey
1991-1994
The Abdus Salam International Centre for Theoretical Physics (ICTP)
1994
Using a combination of perturbation theory and quantum Monte Carlo, we elucidate the behavior single-particle Green's function local spin-spin correlation near Mott transition in infinite dimensional Hubbard model at half filling. The has three fixed points: Fermi liquid phase qualitatively described by Brinkman-Rice picture, insulating resembling Hubbard's solution, an unstable point which connects two.
We discuss the Mott-Hubbard transition in light of Hubbard model infinite dimensions with special emphasis on finite-temperature aspects problem. demonstrate that Mott at finite temperatures has a first-order character. determine region where metallic and insulating solutions coexist using second-order perturbation theory we draw phase diagram half filling semicircular density states. lessons learned from present treatment connection to other approximation schemes experiments transition-metal oxides.
By constructing a complete set of coherent states that forbids double occupancy, we introduce quantum Monte Carlo algorithm for simulation fermion models with constraint, including the infinite-U Hubbard model and t-J model. Application to on 4\ifmmode\times\else\texttimes\fi{}4 6\ifmmode\times\else\texttimes\fi{}6 lattices provides new physical results thermodynamics model, especially temperature doping dependence Nagaoka state.
Hylurgus ligniperda, an invasive species originating from Eurasia, is now a major forestry quarantine pest worldwide. In recent years, it has caused significant damage in China. While traps have been effective monitoring and controlling pests, manual inspections are labor-intensive require expertise insect classification. To address this, we applied two-stage cascade convolutional neural network, YOLOX-MobileNetV2 (YOLOX-Mnet), for identifying H. ligniperda other pests captured traps. This...
Abstract Recent breakthroughs in VR technologies, especially economic headsets and massive smartphones are creating a fast‐growing demand for 3D immersive content. 360 videos record surrounding environment every direction give users fully experience. Thanks to ton of cameras that launched the past years, video content creation is exploding becoming new standard digital industry. When ERP CMP perhaps most prevalent projection packing layout storing videos, they have severe distortion,...
A face recognition system has been developed and demonstrated at the Rutgers University Center for Computer Aids Industrial Productivity. The uses a preliminary data reduction step, gray scale projection, fast transform technique to greatly reduce computational complexity of problem and, consequently, cost high-speed implementation. decision function is new, extremely cost-effective neural network, Mammone/Sankar Neural Tree Network. This paper examines use projection in detail, demonstrates...
A face recognition system has been developed and demonstrated at the Rutgers University Center for Computer Aids Industrial Productivity. The uses a preliminary data reduction step. gray scale projections, fast transform technique to greatly reduce computational complexity of problem and, consequently, cost high-speed implementation. decision function is few, extremely cost-effective neural network, Mammone/Sankar Neural Tree Network. This network can be trained re-trained rapidly on image...
Using path integral quantization in the subspace that forbids double occupancy, we introduce a quantum Monte Carlo algorithm for simulation of fermion models with constraint. The can be applied to class lattice models, including infinite-U Hubbard model and t - J model.
We present an asymptotically exact solution of the $\infty-d$ Hubbard model at a special interaction strength $U_T$ corresponding to strong-coupling Fermi-liquid fixed point. This is intimately related Toulouse limit single-impurity Kondo and symmetric Anderson in its limit.
Green's-function updating is introduced in the recently developed quantum Monte Carlo algorithm for constrained fermions, thus making computing time scale same way as ordinary algorithm. Stabilization routines are also implemented, with limited success lowering down simulation temperature.