This paper investigates the potential of near-field localization using widely-spaced multi-subarrays (WSMSs) and analyzing the corresponding angle and range Cram\'er-Rao bounds (CRBs). By employing the Riemann sum, closed-form CRB expressions are derived for the spherical wavefront-based WSMS (SW-WSMS). We find that the CRBs can be characterized by the angular span formed by the line connecting the array's two ends to the target, and the different WSMSs with same angular spans but different number of subarrays have identical normalized CRBs. We provide a theoretical proof that, in certain scenarios, the CRB of WSMSs is smaller than that of uniform arrays. We further yield the closed-form CRBs for the hybrid spherical and planar wavefront-based WSMS (HSPW-WSMS), and its components can be seen as decompositions of the parameters from the CRBs for the SW-WSMS. Simulations are conducted to validate the accuracy of the derived closed-form CRBs and provide further insights into various system characteristics. Basically, this paper underscores the high resolution of utilizing WSMS for localization, reinforces the validity of adopting the HSPW assumption, and, considering its applications in communications, indicates a promising outlook for integrated sensing and communications based on HSPW-WSMSs.
The human brain exhibits remarkable abilities in integrating temporally distant sensory inputs for decision-making. However, existing brain-inspired spiking neural networks (SNNs) have struggled to match their biological counterpart in modeling long-term temporal relationships. To address this problem, this paper presents a novel Contextual Embedding Leaky Integrate-and-Fire (CE-LIF) spiking neuron model. Specifically, the CE-LIF model incorporates a meticulously designed contextual embedding component into the adaptive neuronal firing threshold, thereby enhancing the memory storage of spiking neurons and facilitating effective sequential modeling. Additionally, theoretical analysis is provided to elucidate how the CE-LIF model enables long-term temporal credit assignment. Remarkably, when compared to state-of-the-art recurrent SNNs, feedforward SNNs comprising the proposed CE-LIF neurons demonstrate superior performance across extensive sequential modeling tasks in terms of classification accuracy, network convergence speed, and memory capacity.
The biological neural systems evolved to adapt to ecological environment for efficiency and effectiveness, wherein neurons with heterogeneous structures and rich dynamics are optimized to accomplish complex cognitive tasks. Most of the current research of biologically inspired spiking neural networks (SNNs) are, however, grounded on a homogeneous neural coding scheme, which limits their overall performance in terms of accuracy, latency, efficiency, and robustness, etc. In this work, we argue that one should holistically design the network architecture to incorporate diverse neuronal functions and neural coding schemes for best performance. As an early attempt in this research direction, we put forward a hybrid neural coding framework that integrates multiple neural coding schemes discovered in neuroscience. We demonstrate that the proposed hybrid coding scheme achieves a comparable accuracy with the state-of-the-art SNNs with homogeneous neural coding on CIFAR-10, CIFAR-100, and Tiny-ImageNet datasets with less than eight time steps and at least 3.90x fewer computations. Furthermore, we demonstrate accurate, rapid, and robust sound source localization on SoClas dataset. This study yields valuable insights into the performance of various hybrid neural coding designs and hold significant implications for designing high performance SNNs.
Real-time emotion-based music arrangement, which aims to transform a given music piece into another one that evokes specific emotional resonance with the user in real-time, holds significant application value in various scenarios, e.g., music therapy, video game soundtracks, and movie scores. However, balancing emotion real-time fit with soft emotion transition is a challenge due to the fine-grained and mutable nature of the target emotion. Existing studies mainly focus on achieving emotion real-time fit, while the issue of soft transition remains understudied, affecting the overall emotional coherence of the music. In this paper, we propose SongDriver2 to address this balance. Specifically, we first recognize the last timestep's music emotion and then fuse it with the current timestep's target input emotion. The fused emotion then serves as the guidance for SongDriver2 to generate the upcoming music based on the input melody data. To adjust music similarity and emotion real-time fit flexibly, we downsample the original melody and feed it into the generation model. Furthermore, we design four music theory features to leverage domain knowledge to enhance emotion information and employ semi-supervised learning to mitigate the subjective bias introduced by manual dataset annotation. According to the evaluation results, SongDriver2 surpasses the state-of-the-art methods in both objective and subjective metrics. These results demonstrate that SongDriver2 achieves real-time fit and soft transitions simultaneously, enhancing the coherence of the generated music.
Real-scanned point clouds are often incomplete due to viewpoint, occlusion, and noise, which hampers 3D geometric modeling and perception. Existing point cloud completion methods tend to generate global shape skeletons and hence lack fine local details. Furthermore, they mostly learn a deterministic partial-to-complete mapping, but overlook structural relations in man-made objects. To tackle these challenges, this paper proposes a variational framework, Variational Relational point Completion Network (VRCNet) with two appealing properties: 1) Probabilistic Modeling. In particular, we propose a dual-path architecture to enable principled probabilistic modeling across partial and complete clouds. One path consumes complete point clouds for reconstruction by learning a point VAE. The other path generates complete shapes for partial point clouds, whose embedded distribution is guided by distribution obtained from the reconstruction path during training. 2) Relational Enhancement. Specifically, we carefully design point self-attention kernel and point selective kernel module to exploit relational point features, which refines local shape details conditioned on the coarse completion. In addition, we contribute multi-view partial point cloud datasets (MVP and MVP-40 dataset) containing over 200,000 high-quality scans, which render partial 3D shapes from 26 uniformly distributed camera poses for each 3D CAD model. Extensive experiments demonstrate that VRCNet outperforms state-of-the-art methods on all standard point cloud completion benchmarks. Notably, VRCNet shows great generalizability and robustness on real-world point cloud scans. Moreover, we can achieve robust 3D classification for partial point clouds with the help of VRCNet, which can highly increase classification accuracy.
Nowadays, the rapid development of photovoltaic(PV) power stations requires increasingly reliable maintenance and fault diagnosis of PV modules in the field. Due to the effectiveness, convolutional neural network (CNN) has been widely used in the existing automatic defect detection of PV cells. However, the parameters of these CNN-based models are very large, which require stringent hardware resources and it is difficult to be applied in actual industrial projects. To solve these problems, we propose a novel lightweight high-performance model for automatic defect detection of PV cells in electroluminescence(EL) images based on neural architecture search and knowledge distillation. To auto-design an effective lightweight model, we introduce neural architecture search to the field of PV cell defect classification for the first time. Since the defect can be any size, we design a proper search structure of network to better exploit the multi-scale characteristic. To improve the overall performance of the searched lightweight model, we further transfer the knowledge learned by the existing pre-trained large-scale model based on knowledge distillation. Different kinds of knowledge are exploited and transferred, including attention information, feature information, logit information and task-oriented information. Experiments have demonstrated that the proposed model achieves the state-of-the-art performance on the public PV cell dataset of EL images under online data augmentation with accuracy of 91.74% and the parameters of 1.85M. The proposed lightweight high-performance model can be easily deployed to the end devices of the actual industrial projects and retain the accuracy.
Adaptive regularization methods that exploit more than the diagonal entries exhibit state of the art performance for many tasks, but can be prohibitive in terms of memory and running time. We find the spectra of the Kronecker-factored gradient covariance matrix in deep learning (DL) training tasks are concentrated on a small leading eigenspace that changes throughout training, motivating a low-rank sketching approach. We describe a generic method for reducing memory and compute requirements of maintaining a matrix preconditioner using the Frequent Directions (FD) sketch. Our technique allows interpolation between resource requirements and the degradation in regret guarantees with rank $k$: in the online convex optimization (OCO) setting over dimension $d$, we match full-matrix $d^2$ memory regret using only $dk$ memory up to additive error in the bottom $d-k$ eigenvalues of the gradient covariance. Further, we show extensions of our work to Shampoo, placing the method on the memory-quality Pareto frontier of several large scale benchmarks.
Tuning optimizer hyperparameters, notably the learning rate to a particular optimization instance, is an important but nonconvex problem. Therefore iterative optimization methods such as hypergradient descent lack global optimality guarantees in general. We propose an online nonstochastic control methodology for mathematical optimization. The choice of hyperparameters for gradient based methods, including the learning rate, momentum parameter and preconditioner, is described as feedback control. The optimal solution to this control problem is shown to encompass preconditioned adaptive gradient methods with varying acceleration and momentum parameters. Although the optimal control problem by itself is nonconvex, we show how recent methods from online nonstochastic control based on convex relaxation can be applied to compete with the best offline solution. This guarantees that in episodic optimization, we converge to the best optimization method in hindsight.
In recent years, aerial swarm technology has developed rapidly. In order to accomplish a fully autonomous aerial swarm, a key technology is decentralized and distributed collaborative SLAM (CSLAM) for aerial swarms, which estimates the relative pose and the consistent global trajectories. In this paper, we propose $D^2$SLAM: a decentralized and distributed ($D^2$) collaborative SLAM algorithm. This algorithm has high local accuracy and global consistency, and the distributed architecture allows it to scale up. $D^2$SLAM covers swarm state estimation in two scenarios: near-field state estimation for high real-time accuracy at close range and far-field state estimation for globally consistent trajectories estimation at the long-range between UAVs. Distributed optimization algorithms are adopted as the backend to achieve the $D^2$ goal. $D^2$SLAM is robust to transient loss of communication, network delays, and other factors. Thanks to the flexible architecture, $D^2$SLAM has the potential of applying in various scenarios.
In recent years, mobile robots are becoming ambitious and deployed in large-scale scenarios. Serving as a high-level understanding of environments, a sparse skeleton graph is beneficial for more efficient global planning. Currently, existing solutions for skeleton graph generation suffer from several major limitations, including poor adaptiveness to different map representations, dependency on robot inspection trajectories and high computational overhead. In this paper, we propose an efficient and flexible algorithm generating a trajectory-independent 3D sparse topological skeleton graph capturing the spatial structure of the free space. In our method, an efficient ray sampling and validating mechanism are adopted to find distinctive free space regions, which contributes to skeleton graph vertices, with traversability between adjacent vertices as edges. A cycle formation scheme is also utilized to maintain skeleton graph compactness. Benchmark comparison with state-of-the-art works demonstrates that our approach generates sparse graphs in a substantially shorter time, giving high-quality global planning paths. Experiments conducted in real-world maps further validate the capability of our method in real-world scenarios. Our method will be made open source to benefit the community.