In autonomous vehicles or robots, point clouds from LiDAR can provide accurate depth information of objects compared with 2D images, but they also suffer a large volume of data, which is inconvenient for data storage or transmission. In this paper, w
e propose a Range image-based Point Cloud Compression method, R-PCC, which can reconstruct the point cloud with uniform or non-uniform accuracy loss. We segment the original large-scale point cloud into small and compact regions for spatial redundancy and salient region classification. Compared with other voxel-based or image-based compression methods, our method can keep and align all points from the original point cloud in the reconstructed point cloud. It can also control the maximum reconstruction error for each point through a quantization module. In the experiments, we prove that our easier FPS-based segmentation method can achieve better performance than instance-based segmentation methods such as DBSCAN. To verify the advantages of our proposed method, we evaluate the reconstruction quality and fidelity for 3D object detection and SLAM, as the downstream tasks. The experimental results show that our elegant framework can achieve 30$times$ compression ratio without affecting downstream tasks, and our non-uniform compression framework shows a great improvement on the downstream tasks compared with the state-of-the-art large-scale point cloud compression methods. Our real-time method is efficient and effective enough to act as a baseline for range image-based point cloud compression. The code is available on https://github.com/StevenWang30/R-PCC.git.
The prompt emission of most gamma-ray bursts (GRBs) typically exhibits a non-thermal Band component. The synchrotron radiation in the popular internal shock model is generally put forward to explain such a non-thermal component. However, the low-ener
gy photon index $alpha sim -1.5$ predicted by the synchrotron radiation is inconsistent with the observed value $alpha sim -1$. Here, we investigate the evolution of a magnetic field during propagation of internal shocks within an ultrarelativistic outflow, and revisit the fast cooling of shock-accelerated electrons via synchrotron radiation for this evolutional magnetic field. We find that the magnetic field is first nearly constant and then decays as $Bpropto t^{-1}$, which leads to a reasonable range of the low-energy photon index, $-3/2 < alpha < -2/3$. In addition, if a rising electron injection rate during a GRB is introduced, we find that $alpha$ reaches $-2/3$ more easily. We thus fit the prompt emission spectra of GRB 080916c and GRB~080825c.
This report includes the original manuscript (pp. 2-40) and the supplementary material (pp. 41-48) of Passive Mechanical Realizations of Bicubic Impedances with No More Than Five Elements for Inerter-Based Control Design.
Tactile sensing plays an important role in robotic perception and manipulation tasks. To overcome the real-world limitations of data collection, simulating tactile response in a virtual environment comes as a desirable direction of robotic research.
In this paper, we propose Elastic Interaction of Particles (EIP) for tactile simulation. Most existing works model the tactile sensor as a rigid multi-body, which is incapable of reflecting the elastic property of the tactile sensor as well as characterizing the fine-grained physical interaction between the two objects. By contrast, EIP models the tactile sensor as a group of coordinated particles, and the elastic property is applied to regulate the deformation of particles during contact. With the tactile simulation by EIP, we further propose a tactile-visual perception network that enables information fusion between tactile data and visual images. The perception network is based on a global-to-local fusion mechanism where multi-scale tactile features are aggregated to the corresponding local region of the visual modality with the guidance of tactile positions and directions. The fusion method exhibits superiority regarding the 3D geometric reconstruction task.
We propose a compact and effective framework to fuse multimodal features at multiple layers in a single network. The framework consists of two innovative fusion schemes. Firstly, unlike existing multimodal methods that necessitate individual encoders
for different modalities, we verify that multimodal features can be learnt within a shared single network by merely maintaining modality-specific batch normalization layers in the encoder, which also enables implicit fusion via joint feature representation learning. Secondly, we propose a bidirectional multi-layer fusion scheme, where multimodal features can be exploited progressively. To take advantage of such scheme, we introduce two asymmetric fusion operations including channel shuffle and pixel shift, which learn different fused features with respect to different fusion directions. These two operations are parameter-free and strengthen the multimodal feature interactions across channels as well as enhance the spatial feature discrimination within channels. We conduct extensive experiments on semantic segmentation and image translation tasks, based on three publicly available datasets covering diverse modalities. Results indicate that our proposed framework is general, compact and is superior to state-of-the-art fusion frameworks.
There is significant interest in learning and optimizing a complex system composed of multiple sub-components, where these components may be agents or autonomous sensors. Among the rich literature on this topic, agent-based and domain-specific simula
tions can capture complex dynamics and subgroup interaction, but optimizing over such simulations can be computationally and algorithmically challenging. Bayesian approaches, such as Gaussian processes (GPs), can be used to learn a computationally tractable approximation to the underlying dynamics but typically neglect the detailed information about subgroups in the complicated system. We attempt to find the best of both worlds by proposing the idea of decomposed feedback, which captures group-based heterogeneity and dynamics. We introduce a novel decomposed GP regression to incorporate the subgroup decomposed feedback. Our modified regression has provably lower variance -- and thus a more accurate posterior -- compared to previous approaches; it also allows us to introduce a decomposed GP-UCB optimization algorithm that leverages subgroup feedback. The Bayesian nature of our method makes the optimization algorithm trackable with a theoretical guarantee on convergence and no-regret property. To demonstrate the wide applicability of this work, we execute our algorithm on two disparate social problems: infectious disease control in a heterogeneous population and allocation of distributed weather sensors. Experimental results show that our new method provides significant improvement compared to the state-of-the-art.
For quantum search via the continuous-time quantum walk, the evolution of the whole system is usually limited in a small subspace. In this paper, we discuss how the symmetries of the graphs are related to the existence of such an invariant subspace,
which also suggests a dimensionality reduction method based on group representation theory. We observe that in the one-dimensional subspace spanned by each desired basis state which assembles the identically evolving original basis states, we always get a trivial representation of the symmetry group. So we could find the desired basis by exploiting the projection operator of the trivial representation. Besides being technical guidance in this type of problem, this discussion also suggests that all the symmetries are used up in the invariant subspace and the asymmetric part of the Hamiltonian is very important for the purpose of quantum search.
In the predict-then-optimize framework, the objective is to train a predictive model, mapping from environment features to parameters of an optimization problem, which maximizes decision quality when the optimization is subsequently solved. Recent wo
rk on decision-focused learning shows that embedding the optimization problem in the training pipeline can improve decision quality and help generalize better to unseen tasks compared to relying on an intermediate loss function for evaluating prediction quality. We study the predict-then-optimize framework in the context of sequential decision problems (formulated as MDPs) that are solved via reinforcement learning. In particular, we are given environment features and a set of trajectories from training MDPs, which we use to train a predictive model that generalizes to unseen test MDPs without trajectories. Two significant computational challenges arise in applying decision-focused learning to MDPs: (i) large state and action spaces make it infeasible for existing techniques to differentiate through MDP problems, and (ii) the high-dimensional policy space, as parameterized by a neural network, makes differentiating through a policy expensive. We resolve the first challenge by sampling provably unbiased derivatives to approximate and differentiate through optimality conditions, and the second challenge by using a low-rank approximation to the high-dimensional sample-based derivatives. We implement both Bellman--based and policy gradient--based decision-focused learning on three different MDP problems with missing parameters, and show that decision-focused learning performs better in generalization to unseen tasks.
A growing body of work in game theory extends the traditional Stackelberg game to settings with one leader and multiple followers who play a Nash equilibrium. Standard approaches for computing equilibria in these games reformulate the followers best
response as constraints in the leaders optimization problem. These reformulation approaches can sometimes be effective, but often get trapped in low-quality solutions when followers objectives are non-linear or non-quadratic. Moreover, these approaches assume a unique equilibrium or a specific equilibrium concept, e.g., optimistic or pessimistic, which is a limiting assumption in many situations. To overcome these limitations, we propose a stochastic gradient descent--based approach, where the leaders strategy is updated by differentiating through the followers best responses. We frame the leaders optimization as a learning problem against followers equilibrium, which allows us to decouple the followers equilibrium constraints from the leaders problem. This approach also addresses cases with multiple equilibria and arbitrary equilibrium selection procedures by back-propagating through a sampled Nash equilibrium. To this end, this paper introduces a novel concept called equilibrium flow to formally characterize the set of equilibrium selection processes where the gradient with respect to a sampled equilibrium is an unbiased estimate of the true gradient. We evaluate our approach experimentally against existing baselines in three Stackelberg problems with multiple followers and find that in each case, our approach is able to achieve higher utility for the leader.
Face recognition has achieved significant progress in deep-learning era due to the ultra-large-scale and well-labeled datasets. However, training on ultra-large-scale datasets is time-consuming and takes up a lot of hardware resource. Therefore,
designing an efficient training approach is crucial and indispensable. The heavy computational and memory costs mainly result from the high dimensionality of the Fully-Connected (FC) layer. Specifically, the dimensionality is determined by the number of face identities, which can be million-level or even more. To this end, we propose a novel training approach for ultra-large-scale face datasets, termed Faster Face Classification (F$^2$C). In F$^2$C, we first define a Gallery Net and a Probe Net that are used to generate identities centers and extract faces features for face recognition, respectively. Gallery Net has the same structure as Probe Net and inherits the parameters from Probe Net with a moving average paradigm. After that, to reduce the training time and hardware costs of the FC layer, we propose a Dynamic Class Pool (DCP) that stores the features from Gallery Net and calculates the inner product (logits) with positive samples (whose identities are in the DCP) in each mini-batch. DCP can be regarded as a substitute for the FC layer but it is far smaller, thus greatly reducing the computational and memory costs. For negative samples (whose identities are not in DCP), we minimize the cosine similarities between negative samples and those in DCP. Then, to improve the update efficiency of DCPs parameters, we design a dual data-loader including identity-based and instance-based loaders to generate a certain of identities and samples in mini-batches.
