arXiv subjects

Yonggang Yang

Publications and source records attributed to Yonggang Yang.

10 recordsLinked to original sources

arXiv2026

Multi-Branch Non-Homogeneous Image Dehazing via Concentration Partitioning and Image Fusion

Existing single image dehazing methods have demonstrated satisfactory performance on homogeneous thin-haze images; however, they often struggle with non-homogeneous hazy images that exhibit spatially varying haze concentrations and abrupt density transitions across different regions. To address this fundamental limitation, we propose a novel multi-branch deep neural network framework, termed Concentration Partitioning and Image Fusion Network (CPIFNet), which decomposes the challenging non-homogeneous dehazing problem into a set of tractable homogeneous sub-problems. Our key insight is that a single non-homogeneous hazy image can be viewed as a composite of multiple local regions, each exhibiting approximately homogeneous haze characteristics. CPIFNet employs a two-stage architecture consisting of an Image Enhancement Network (IENet) stage and an Image Fusion Network (IFNet) stage. In the first stage, multiple IENet branches are independently trained on homogeneous haze datasets of different concentration levels, producing enhancement models that excel at restoring regions matching their respective haze densities. In the second stage, the IFNet intelligently aggregates the advantageous regions from all enhancement outputs through deep feature stacking and merging, yielding a unified high-quality dehazed result. Furthermore, we introduce a comprehensive loss function incorporating reconstruction, perceptual, structural, and color losses to jointly supervise both stages.

cs.CV↗

arXiv2026

Selective Attention-Based Network for Robust Infrared Small Target Detection

Infrared small target detection (IRSTD) plays a pivotal role in a broad spectrum of mission-critical applications, including maritime surveillance, military search and rescue, early warning systems, and precision-guided strikes, all of which demand the precise identification of dim, sub-pixel targets amid highly cluttered infrared backgrounds. Despite significant progress driven by deep learning methods, fundamental challenges persist: infrared small targets occupy extremely limited spatial extents (often only a few pixels), exhibit low signal-to-clutter ratios, and are easily confused with structurally complex backgrounds that frequently induce false alarms. Existing encoder-decoder architectures suffer from two key limitations - an information bottleneck in early convolutional stages that undermines fine-grained target perception, and static skip connections that lack the dynamic adaptability required to discriminate between genuine targets and pseudo-target regions. To address these challenges, we propose SANet, a Selective Attention-based Network built upon the classical U-Net framework and augmented with two novel components: (1) a \emph{Dual-path Semantic-aware Module} (DSM) that integrates standard convolutions for local spatial detail preservation with pinwheel-shaped convolutions for expanded, direction-sensitive receptive fields, followed by a Convolutional Block Attention Module (CBAM) for fine-grained spatial-channel feature recalibration; and (2) a \emph{Selective Attention Fusion Module} (SAFM) that replaces conventional static skip connections with a spatially adaptive, learnable weighting mechanism to perform context-aware, cross-scale feature fusion.

cs.CV↗

arXiv2020

Super interference fringes of two-photon photoluminescence in individual Au nanoparticles: the critical role of the intermediate state

The interaction between light and metal nanoparticles enables investigations of microscopic phenomena on nanometer length and ultrashort time scales, benefiting from strong confinement and enhancement of the optical field. However, the ultrafast dynamics of these nanoparticles are primarily investigated by multiphoton photoluminescence on picoseconds or photoemission on femtoseconds independently. Here, we presented two-photon photoluminescence (TPPL) measurements on individual Au nanobipyramids (AuNP) to reveal their ultrafast dynamics by two-pulse excitation on a global time scale ranging from sub-femtosecond to tens of picoseconds. Two-orders-of-magnitude photoluminescence enhancement, namely super interference fringes, has been demonstrated on tens of femtoseconds. Power-dependent measurements uncovered the transform of the nonlinearity from 1 to 2 when the interpulse delay varied from tens of femtoseconds to tens of picoseconds. We proved that the real intermediate state plays a critical role in the observed phenomena, supported by numerical simulations with a three eigenstates model and further experiments on Au nanospheres with different diameters. The crucial parameters, including the dephasing time, the radiative rate, and the coupling between different states, have been estimated using numerical simulations. Our results provide insight into the role of intermediate states in the ultrafast dynamics of noble metal nanoparticles. The giant photoluminescence in super interference fringes enables potential practical applications in imaging, sensing, and nanophotonics.

physics.optics↗

arXiv2015

Dipolar Rydberg-atom gas prepared by adiabatic passage through an avoided crossing

The passage of cold cesium 49S$_{1/2}$ Rydberg atoms through an electric-field-induced multi-level avoided crossing with nearby hydrogen-like Rydberg levels is employed to prepare a cold, dipolar Rydberg atom gas. When the electric field is ramped through the avoided crossing on time scales on the order of 100~ns or slower, the 49S$_{1/2}$ population adiabatically transitions into high-\emph{l} Rydberg Stark states. The adiabatic state transformation results in a cold gas of Rydberg atoms with large electric dipole moments. After a waiting time of about $1~μ$s and at sufficient atom density, the adiabatically transformed highly dipolar atoms become undetectable, enabling us to discern adiabatic from diabatic passage behavior through the avoided crossing. We attribute the state-selectivity to $m$-mixing collisions between the dipolar atoms. The data interpretation is supported by numerical simulations of the passage dynamics and of binary $m$-mixing collisions.

physics.atom-ph↗

arXiv2014

Maximum tunneling velocities in symmetric double well potentials

We consider coherent tunneling of one-dimensional model systems in non-cyclic or cyclic symmetric double well potentials. Generic potentials are constructed which allow for analytical estimates of the quantum dynamics in the non-relativistic deep tunneling regime, in terms of the tunneling distance, barrier height and mass (or moment of inertia). For cyclic systems, the results may be scaled to agree well with periodic potentials for which semi-analytical results in terms of Mathieu functions exist. Starting from a wavepacket which is initially localized in one of the potential wells, the subsequent periodic tunneling is associated with tunneling velocities. These velocities (or angular velocities) are evaluated as the ratio of the flux densities versus the probability densities. The maximum velocities are found under the top of the barrier where they scale as the square root of the ratio of barrier height and mass (or moment of inertia), independent of the tunneling distance. They are applied exemplarily to several prototypical molecular models of non-cyclic and cyclic tunneling, including ammonia inversion, Cope rearrangment of semibullvalene, torsions of molecular fragments, and rotational tunneling in strong laser fields. Typical maximum velocities and angular velocities are in the order of a few km/s and from 10 to 100 THz for our non-cyclic and cyclic systems, respectively, much faster than time-averaged velocities. Even for the more extreme case of an electron tunneling through a barrier of height of one Hartree, the velocity is only about one percent of the speed of light. Estimates of the corresponding time scales for passing through {the narrow domain just} below the potential barrier are in the domain from 2 to 40 fs, much shorter than the tunneling times.

physics.chem-ph↗

arXiv2013

Optimization and analysis of experimental parameters for polarization gradient cooling in optical molasses

We systematically investigate the dependence of the temperature of cold cesium atoms of polarization gradient cooling (PGC) in optical molasses on experimental parameters, which contain changing modes of cooling laser, PGC interaction time, cooling laser frequency and its intensity. The SR mode of cooling laser, that means the cooling laser frequency is changed with step mode and cooling laser intensity is changed with ramp mode, is found to be the best for PGC comparing with other SS, RS, and RR modes. We introduce a statistical explanation and an exponential decay function to explain the variation of cold atomic temperature on time. The heating effect is observed when the cooling laser intensity is lower than the saturation intensity of cold atoms. After optimization, the lowest temperature of cold cesium atoms is observed to be about 4uK with the number of 2x10^9, a density of 1x10^11/cm^3 and the phase space density of 4.4x10^(-5). The optimization process and analysis of controllable experimental parameters are also meaningful for other cold atomic systems.

physics.atom-ph↗

arXiv2012

Ultralong Rydberg Cs$_2$ Molecules Investigated by Combined ab initio Calculations and Perturbation Theory

Vibrational properties of ultralong Rydberg Cs$_2$ molecules are investigated on corresponding potential curves obtained by perturbation theory. The Rydberg Cs$_2$ molecules are associated by a Rydberg Cs($nS/nP$) atom $(n=30-70)$ and a ground state Cs($6s$) atom. The starting point for the perturbation treatment of corresponding Rydberg molecular potential curves is to generate accurate atomic Rydberg states from realistic {\it ab initio} effective core potential. The calculated results have similar characteristics with available experimental and theoretical investigations on Rydberg Rb$_2$ molecules. And this is the first time that Rydberg molecules are studied at the {\it ab initio} level.

physics.atom-ph↗

arXiv2011

Path Integral Approach to the Calculation of Reaction Rates for a Reaction Coordinate Coupled to a Dual Harmonic Bath

We present a new method for the numerical calculation of canonical reaction rate constants in complex molecular systems, which is based on a path integral formulation of the flux-flux correlation function. Central is the partitioning of the total system into a relevant part coupled to a dual bath. The latter consists of two parts: First, a set of strongly coupled harmonic modes, describing, for example, intramolecular degrees of freedom. They are treated on the basis of a reaction surface Hamiltonian approach. Second, a set of bath modes mimicking an unspecific environment modeled by means of a continuous spectral density. After deriving a set of general equations expressing the canonical rate constant in terms of appropriate influence functionals, several approximations are introduced to provide an efficient numerical implementation. Results for an initial application to the H-transfer in 6-Aminofulvene-1-aldimine are discussed.

physics.chem-ph↗

arXiv2008

A Concise Method for Kinetic Energy Quantisation

We present a straightforward method for obtaining exact classical and quantum molecular Hamiltonians in terms of arbitrary coordinates. As compared to other approaches the resulting expression are rather compact, the physical meaning of each quantity is quite transparent, and in some cases the calculation effort will be greatly reduced. We also investigate systems with constrains to find the suggested method to be applicable in contrast to most conventional approaches to kinetic energy operators which cannot directly be applied to constrained systems. Two examples are discussed in detail, the monohydrated hydroxide anion and the protonated ammonia dimer.

physics.chem-ph↗

arXiv2008

A full-dimensional quantum dynamical study of the vibrational ground state of $H_3O_2^-$ and its isotopomers

We investigated the effect of deuteration on the vibrational ground state of the hydrated hydroxide anion using a nine-dimensional quantum dynamical model for the case of J=0. The propagation of the nuclear wave function has been performed with the multi-configuration time-dependent Hartree method which yielded zero-point energies for the normal and fully deuterated species in quantitative agreement with previous diffusion Monte Carlo calculations. According to the zero-point energy the isotopomers having the hydrogen atom in the bridging position are more stable by about 1 kJ/mol as compared to the deuterium case. This holds irrespective of the deuteration state of the two OH groups. We also report the secondary geometric H/D isotope effect on the O--O distance which amounts to an elongation of about 0.005 A for the symmetric isotopomers and 0.009 A in the asymmetric case. Finally, we explore the isotopomer sensitivity of the ground state tunneling splitting due to the torsional motion of the two OH groups.

physics.chem-ph↗