Search

Traffic prediction is a challenging spatio-temporal forecasting problem that involves highly complex spatio-temporal correlations. This paper proposes a Multi-level Multi-view Augmented Spatio-temporal Transformer (LVSTformer) for traffic prediction. The model aims to capture spatial dependencies from three different levels: local geographic, global semantic, and pivotal nodes, along with long- and short-term temporal dependencies. Specifically, we design three spatial augmented views to delve into the spatial information from the perspectives of local, global, and pivotal nodes. By combining three spatial augmented views with three parallel spatial self-attention mechanisms, the model can comprehensively captures spatial dependencies at different levels. We design a gated temporal self-attention mechanism to effectively capture long- and short-term temporal dependencies. Furthermore, a spatio-temporal context broadcasting module is introduced between two spatio-temporal layers to ensure a well-distributed allocation of attention scores, alleviating overfitting and information loss, and enhancing the generalization ability and robustness of the model. A comprehensive set of experiments is conducted on six well-known traffic benchmarks, the experimental results demonstrate that LVSTformer achieves state-of-the-art performance compared to competing baselines, with the maximum improvement reaching up to 4.32%.

3D Gaussian Splatting showcases notable advancements in photo-realistic and real-time novel view synthesis. However, it faces challenges in modeling mirror reflections, which exhibit substantial appearance variations from different viewpoints. To tackle this problem, we present MirrorGaussian, the first method for mirror scene reconstruction with real-time rendering based on 3D Gaussian Splatting. The key insight is grounded on the mirror symmetry between the real-world space and the virtual mirror space. We introduce an intuitive dual-rendering strategy that enables differentiable rasterization of both the real-world 3D Gaussians and the mirrored counterpart obtained by reflecting the former about the mirror plane. All 3D Gaussians are jointly optimized with the mirror plane in an end-to-end framework. MirrorGaussian achieves high-quality and real-time rendering in scenes with mirrors, empowering scene editing like adding new mirrors and objects. Comprehensive experiments on multiple datasets demonstrate that our approach significantly outperforms existing methods, achieving state-of-the-art results. Project page: https://mirror-gaussian.github.io/.

Equilibrium Propagation (EP) is a biologically plausible local learning algorithm initially developed for convergent recurrent neural networks (RNNs), where weight updates rely solely on the connecting neuron states across two phases. The gradient calculations in EP have been shown to approximate the gradients computed by Backpropagation Through Time (BPTT) when an infinitesimally small nudge factor is used. This property makes EP a powerful candidate for training Spiking Neural Networks (SNNs), which are commonly trained by BPTT. However, in the spiking domain, previous studies on EP have been limited to architectures involving few linear layers. In this work, for the first time we provide a formulation for training convolutional spiking convergent RNNs using EP, bridging the gap between spiking and non-spiking convergent RNNs. We demonstrate that for spiking convergent RNNs, there is a mismatch in the maximum pooling and its inverse operation, leading to inaccurate gradient estimation in EP. Substituting this with average pooling resolves this issue and enables accurate gradient estimation for spiking convergent RNNs. We also highlight the memory efficiency of EP compared to BPTT. In the regime of SNNs trained by EP, our experimental results indicate state-of-the-art performance on the MNIST and FashionMNIST datasets, with test errors of 0.97% and 8.89%, respectively. These results are comparable to those of convergent RNNs and SNNs trained by BPTT. These findings underscore EP as an optimal choice for on-chip training and a biologically-plausible method for computing error gradients.

Existing NeRF-based methods for large scene reconstruction often have limitations in visual quality and rendering speed. While the recent 3D Gaussian Splatting works well on small-scale and object-centric scenes, scaling it up to large scenes poses challenges due to limited video memory, long optimization time, and noticeable appearance variations. To address these challenges, we present VastGaussian, the first method for high-quality reconstruction and real-time rendering on large scenes based on 3D Gaussian Splatting. We propose a progressive partitioning strategy to divide a large scene into multiple cells, where the training cameras and point cloud are properly distributed with an airspace-aware visibility criterion. These cells are merged into a complete scene after parallel optimization. We also introduce decoupled appearance modeling into the optimization process to reduce appearance variations in the rendered images. Our approach outperforms existing NeRF-based methods and achieves state-of-the-art results on multiple large scene datasets, enabling fast optimization and high-fidelity real-time rendering., Comment: Accepted to CVPR 2024. Project website: https://vastgaussian.github.io

Spiking Neural Networks (SNNs), providing more realistic neuronal dynamics, have shown to achieve performance comparable to Artificial Neural Networks (ANNs) in several machine learning tasks. Information is processed as spikes within SNNs in an event-based mechanism that significantly reduces energy consumption. However, training SNNs is challenging due to the non-differentiable nature of the spiking mechanism. Traditional approaches, such as Backpropagation Through Time (BPTT), have shown effectiveness but comes with additional computational and memory costs and are biologically implausible. In contrast, recent works propose alternative learning methods with varying degrees of locality, demonstrating success in classification tasks. In this work, we show that these methods share similarities during the training process, while they present a trade-off between biological plausibility and performance. Further, this research examines the implicitly recurrent nature of SNNs and investigates the influence of addition of explicit recurrence to SNNs. We experimentally prove that the addition of explicit recurrent weights enhances the robustness of SNNs. We also investigate the performance of local learning methods under gradient and non-gradient based adversarial attacks.

The ubiquitous existence of microbial communities marks the importance of understanding how species interact within the community to coexist and their spatial organization. We study a two-species mutualistic cross-feeding model through a stochastic cellular automaton on a square lattice using kinetic Monte Carlo simulation. Our model encapsulates the essential dynamic processes such as cell growth, and nutrient excretion, diffusion and uptake. Focusing on the interplay among nutrient diffusion and individual cell division, we discover three general classes of colony morphology: co-existing sectors, co-existing spirals, and engulfment. When the cross-feeding nutrient is widely available, either through high excretion or fast diffusion, a stable circular colony with alternating species sector emerges. When the consumer cells rely on being spatially close to the producers, we observe a stable spiral. We also see one species being engulfed by the other when species interfaces merge due to stochastic fluctuation. By tuning the diffusion rate and the growth rate, we are able to gain quantitative insights into the structures of the sectors and the spirals., Comment: 17 pages 17 figures

Pruning techniques have been successfully used in neural networks to trade accuracy for sparsity. However, the impact of network pruning is not uniform: prior work has shown that the recall for underrepresented classes in a dataset may be more negatively affected. In this work, we study such relative distortions in recall by hypothesizing an intensification effect that is inherent to the model. Namely, that pruning makes recall relatively worse for a class with recall below accuracy and, conversely, that it makes recall relatively better for a class with recall above accuracy. In addition, we propose a new pruning algorithm aimed at attenuating such effect. Through statistical analysis, we have observed that intensification is less severe with our algorithm but nevertheless more pronounced with relatively more difficult tasks, less complex models, and higher pruning ratios. More surprisingly, we conversely observe a de-intensification effect with lower pruning ratios, which indicates that moderate pruning may have a corrective effect to such distortions., Comment: NeurIPS 2022

The statistical study of the Coronal Mass Ejections (CMEs) is a hot topic in solar physics. To further reveal the temporal and spatial behaviors of the CMEs at different latitudes and heights, we analyzed the correlation and phase relationships between the occurrence rate of CMEs, the Coronal Brightness Index (CBI), and the 10.7-cm solar radio flux (F10.7). We found that the occurrence rate of the CMEs correlates with CBI relatively stronger at high latitudes (>=60) than at low latitudes (<=50). At low latitudes, the occurrence rate of the CMEs correlates relatively weaker with CBI than F10.7. There is a relatively stronger correlation relationship between CMEs, F10.7, and CBI during Solar Cycle 24(SC24) than Solar Cycle 23 (SC23). During SC23, the high-latitude CME occurrence rate lags behind F10.7 by three months, and during SC24, the low-latitude CME occurrence rate leads to the low-latitude CBI by one month. The correlation coefficient values turn out to be larger when the very faint CMEsare removed from the samples of the CDAW catalog. Based on our results, we may speculate that the source regions of the high/low-latitude CMEs may vary in height, and the process of magnetic energy accumulation and dissipation is from the lower to the upper atmosphere of the Sun. The temporal offsets between different indicators could help us better understand the physical processes responsible for the solar-terrestrial interactions., Comment: 11 pages, 7 figures, 2 tables. Accepted by APJ

The manifestation of skyrmions in the Mott-insulator Cu$_2$OSeO$_3$ originates from a delicate balance between magnetic and electronic energy scales. As a result of these intertwined couplings, the two symmetry-inequivalent magnetic ions, Cu-I and Cu-II, bond into a spin S=1 entangled tetrahedron. However, conceptualizing the unconventional properties of this material and the energy of the competing interactions is a challenging task due the complexity of this system. Here we combine X-ray Absorption Spectroscopy and Resonant Inelastic X-ray Scattering to uncover the electronic and magnetic excitations of Cu$_2$OSeO$_3$ with site-specificity. We quantify the energies of the 3d crystal-field splitting for both Cu-I and Cu-II, fundamental to optimize model Hamiltonians. Additionally, we unveil a site-specific magnetic mode, indicating that individual spin character is preserved within the entangled-tetrahedron picture. Our results thus provide experimental constraint for validating theories that describe the interactions of Cu$_2$OSeO$_3$, highlighting the site-selective capabilities of resonant spectroscopies

5d iridates have shown vast emergent phenomena due to a strong interplay among its lattice, charge and spin degrees of freedom, because of which the potential in spintronic application of the thin-film form is highly leveraged. Here we have epitaxially stabilized perovskite SrIr$_{0.8}$Sn$_{0.2}$O$_3$ on [001] SrTiO$_3$ substrates through pulsed laser deposition and systematically characterized the structural, electronic and magnetic properties. Physical properties measurements unravel an insulating ground state with a weak ferromagnetism in the compressively strained epitaxial film. The octahedral rotation pattern is identified by synchrotron x-ray diffraction, resolving a mix of $a^+b^-c^-$ and $a^-b^+c^-$ domains. X-ray magnetic resonant scattering directly demonstrates a G-type antiferromagnetic structure of the magnetic order and the spin canting nature of the weak ferromagnetism., Comment: Accepted by Physical Review Material