Search

In this work, the heat vortexes in two-dimensional porous or ribbon structures are investigated based on the phonon Boltzmann transport equation (BTE) under the Callaway model. First, the separate thermal effects of normal (N) scattering and resistive (R) scattering are investigated with frequency-independent assumptions. And then the heat vortexes in graphene are studied as a specific example. It is found that the heat vortexes can appear in both ballistic (rare R/N scattering) and hydrodynamic (N scattering dominates) regimes but disappear in the diffusive (R scattering dominates) regime. As long as there is not sufficient R scattering, the heat vortexes can appear in present simulations.

Recently, context reasoning using image regions beyond local convolution has shown great potential for scene parsing. In this work, we explore how to incorporate the linguistic knowledge to promote context reasoning over image regions by proposing a Graph Interaction unit (GI unit) and a Semantic Context Loss (SC-loss). The GI unit is capable of enhancing feature representations of convolution networks over high-level semantics and learning the semantic coherency adaptively to each sample. Specifically, the dataset-based linguistic knowledge is first incorporated in the GI unit to promote context reasoning over the visual graph, then the evolved representations of the visual graph are mapped to each local representation to enhance the discriminated capability for scene parsing. GI unit is further improved by the SC-loss to enhance the semantic representations over the exemplar-based semantic graph. We perform full ablation studies to demonstrate the effectiveness of each component in our approach. Particularly, the proposed GINet outperforms the state-of-the-art approaches on the popular benchmarks, including Pascal-Context and COCO Stuff., Comment: Accepted by ECCV 2020

Thermal rectification which is a diode-like behavior of heat flux has been studied over a long time. However, a universal and systematic physical description is still lacking. In this letter, a perturbation theory of thermal rectification is developed, which provides an analytical formula of the thermal rectification ratio. It reveals the linear relationship between the thermal rectification ratio and temperature difference. Furthermore, the size-dependence of the thermal rectification relies on the specific form of the thermal conductivity. In addition, several experimental and numerical observations in previous literatures are well explained. This theory can be applicable to any system in which a differentiable effective thermal conductivity can be derived, and is helpful to unveil general principle for thermal rectification., Comment: 5 pages, 2 figures, 50 references, 1 supplemental material

Ship detection using high-resolution remote sensing images is an important task, which contribute to sea surface regulation. The complex background and special visual angle make ship detection relies in high quality datasets to a certain extent. However, there is few works on giving both precise classification and accurate location of ships in existing ship detection datasets. To further promote the research of ship detection, we introduced a new fine-grained ship detection datasets, which is named as FGSD. The dataset collects high-resolution remote sensing images that containing ship samples from multiple large ports around the world. Ship samples were fine categorized and annotated with both horizontal and rotating bounding boxes. To further detailed the information of the dataset, we put forward a new representation method of ships' orientation. For future research, the dock as a new class was annotated in the dataset. Besides, rich information of images were provided in FGSD, including the source port, resolution and corresponding GoogleEarth' s resolution level of each image. As far as we know, FGSD is the most comprehensive ship detection dataset currently and it'll be available soon. Some baselines for FGSD are also provided in this paper.

Human parsing is an essential branch of semantic segmentation, which is a fine-grained semantic segmentation task to identify the constituent parts of human. The challenge of human parsing is to extract effective semantic features to resolve deformation and multi-scale variations. In this work, we proposed an end-to-end model called C-DLinkNet based on LinkNet, which contains a new module named Smooth Module to combine the multi-level features in Decoder part. C-DLinkNet is capable of producing competitive parsing performance compared with the state-of-the-art methods with smaller input sizes and no additional information, i.e., achiving mIoU=53.05 on the validation set of LIP dataset.

Recent neural sequence to sequence models have provided feasible solutions for abstractive summarization. However, such models are still hard to tackle long text dependency in the summarization task. A high-quality summarization system usually depends on strong encoder which can refine important information from long input texts so that the decoder can generate salient summaries from the encoder's memory. In this paper, we propose an aggregation mechanism based on the Transformer model to address the challenge of long text representation. Our model can review history information to make encoder hold more memory capacity. Empirically, we apply our aggregation mechanism to the Transformer model and experiment on CNN/DailyMail dataset to achieve higher quality summaries compared to several strong baseline models on the ROUGE metrics.

The radial thermal rectification in the concentric silicon ring from ballistic to diffusive regime is investigated based on the phonon Boltzmann transport equation. In the ballistic and diffusive limits, the analytical solutions prove that there is no thermal rectification. In the ballistic-diffusive regime, the heat flux prefers to flow from the inner boundary to the outer boundary. Furthermore, as the characteristic length (the distance between two circular boundaries) increases from tens of nanometers to tens of microns, the thermal rectification ratio enhances first and then fades away gradually. It attributes to that as the direction of the temperature gradient changes, the average phonon mean free path changes. The difference of the average phonon mean free path finally leads to the change of the heat flux or thermal conductivity. As the temperature decreases, the maximum thermal rectification ratio decreases. In addition, as the radius ratio between the inner and outer boundary increases, the thermal rectification ratio decreases for a given characteristic length., Comment: 17 pages, 10 figures, 53 references

Hotspot is a ubiquitous phenomenon in microdevices/chips. In homogeneous nanoscale graphene disk with a hotspot, a graded thermal conductivity is observed previously even when the system size is fixed. However, the underlying physical mechanism is not clear. In this work, the hotspots in homogeneous 2D disk/3D ball and graphene disk are studied based on phonon Boltzmann transport equation. The mechanisms of phonon scattering are analyzed. It is found that for a system with fixed size, the graded thermal conductivity is predictable as long as there is not sufficient phonon scattering, which is independent on material properties, dimensions or system size. This work may shed light on both theoretical and experimental studies on heat dissipation of microelectronics., Comment: 5 pages, 3 figures, 48 references

In this paper, a fast synthetic iterative scheme is developed to accelerate convergence for the implicit DOM based on the stationary phonon BTE. The key innovative point of the present scheme is the introduction of the macroscopic synthetic diffusion equation for the temperature, which is obtained from the zero- and first-order moment equations of the phonon BTE. The synthetic diffusion equation, which is asymptomatically preserving to the Fourier's heat conduction equation in the diffusive regime, contains a term related to the Fourier's law and a term determined by the second-order moment of the distribution function that reflects the non-Fourier heat transfer. The mesoscopic kinetic equation and macroscopic diffusion equations are tightly coupled together, because the diffusion equation provides the temperature for the BTE, while the BTE provides the high-order moment to the diffusion equation to describe the non-Fourier heat transfer. This synthetic iterative scheme strengthens the coupling of all phonons in the phase space to facilitate the fast convergence from the diffusive to ballistic regimes. Typical numerical tests in one-, two-, and three-dimensional problems demonstrate that our scheme can describe the multiscale heat transfer problems accurately and efficiently. For all test cases convergence is reached within one hundred iteration steps, which is one to three orders of magnitude faster than the traditional implicit DOM in the near-diffusive regime., Comment: 23 pages, 17 figures, 61 references

In this paper, a finite-volume discrete unified gas kinetic scheme (DUGKS) based on the non-gray phonon transport model is developed for multiscale heat transfer problem with arbitrary temperature difference. Under large temperature difference, the phonon Boltzmann transport equation (BTE) is essentially multiscale, not only in the frequency space, but also in the spatial space. In order to realize the efficient coupling of the multiscale phonon transport, the phonon scattering and advection are coupled together in the present scheme on the reconstruction of the distribution function at the cell interface. The Newtonian method is adopted to solve the nonlinear scattering term for the update of the temperature at both the cell center and interface. In addition, the energy at the cell center is updated by a macroscopic equation instead of taking the moment of the distribution function, which enhances the numerical conservation. Numerical results of the cross-plane heat transfer prove that the present scheme can describe the multiscale heat transfer phenomena accurately with arbitrary temperature difference in a wide range. In the diffusive regime, even if the time step is larger than the relaxation time, the present scheme can capture the transient thermal transport process accurately. Compared to that under small temperature differences, as the temperature difference increases, the variation of the temperature distribution behaves quite differently and the average temperature in the domain increases in the ballistic regime but decreases in the diffusive regime., Comment: 19 pages, 16 figures, 1 table, 47 references