Search

As a broadly applied technique in numerous optimization problems, recently, local search has been employed to solve Pseudo-Boolean Optimization (PBO) problem. A representative local search solver for PBO is LSPBO. In this paper, firstly, we improve LSPBO by a dynamic scoring mechanism, which dynamically strikes a balance between score on hard constraints and score on the objective function. Moreover, on top of this improved LSPBO , we develop the first parallel local search PBO solver. The main idea is to share good solutions among different threads to guide the search, by maintaining a pool of feasible solutions. For evaluating solutions when updating the pool, we propose a function that considers both the solution quality and the diversity of the pool. Furthermore, we calculate the polarity density in the pool to enhance the scoring function of local search. Our empirical experiments show clear benefits of the proposed parallel approach, making it competitive with the parallel version of the famous commercial solver Gurobi., Comment: 17 pages,2 figures, to be published in The 30th International Conference on Principles and Practice of Constraint Programming

Offline reinforcement learning (RL) is crucial for real-world applications where exploration can be costly or unsafe. However, offline learned policies are often suboptimal, and further online fine-tuning is required. In this paper, we tackle the fundamental dilemma of offline-to-online fine-tuning: if the agent remains pessimistic, it may fail to learn a better policy, while if it becomes optimistic directly, performance may suffer from a sudden drop. We show that Bayesian design principles are crucial in solving such a dilemma. Instead of adopting optimistic or pessimistic policies, the agent should act in a way that matches its belief in optimal policies. Such a probability-matching agent can avoid a sudden performance drop while still being guaranteed to find the optimal policy. Based on our theoretical findings, we introduce a novel algorithm that outperforms existing methods on various benchmarks, demonstrating the efficacy of our approach. Overall, the proposed approach provides a new perspective on offline-to-online RL that has the potential to enable more effective learning from offline data., Comment: Forty-first International Conference on Machine Learning (ICML), 2024

Order and disorder constitute two fundamental and opposite themes in condensed matter physics and materials science. Crystals are considered the epitome of order characterized by long-range translational order. The discovery of quasicrystals, with no periodicity but rotational symmetries forbidden for crystals, leads to a paradigm shift in solid-state physics. Moving one step forward, it is intriguing to ask whether ordered matter exists without apparent symmetry breaking. The same question may arise in the pursuit of how ordered amorphous (noncrystalline) solids can be. Here we report the finding of ideal noncrystals in two dimensions, which are disordered in the conventional sense without Bragg peaks but highly ordered according to the steric order. We find that such ideal noncrystals have vibrational modes the same as phonons following the Debye law. The elastic responses are fully affine, which is again characteristic of crystals, and the spatial fluctuations of local volume fractions approach hyperuniformity. Therefore, ideal noncrystals represent an anomalous form of matter with a mixed nature of noncrystalline structure but crystal-like properties. Since such states are found to be thermodynamically favorable, we identify them as a possible new class of ordered matter without apparent broken symmetry. Our results thus extend the scope of the ordered state of matter and may impact the understanding of entropy-driving ordering also in generic amorphous materials.

Polarization engineering has been proven to enhance the capabilities of light manipulation and thus facilitate the development of integrated photonic devices. In this study, we introduce a polarization rotator based on gradiently-twisted $\alpha$-MoO$_3$ thin film, that works for the mid infrared range and functions in a transmission mode. To be specific, the proposed device is constructed by gradiently-twisted $\alpha$-MoO$_3$ multilayers with a subwavelength thickness of only 5 microns, namely, one-third of the working wavelength. Our analytical calculation demonstrates the efficacy of this subwavelength thin film rotator in converting a linearly polarized wave into its orthogonal counterpart, thanks to its chiral nature. The twisted $\alpha$-MoO$_3$ multilayers exhibit the capability to significantly manipulate dispersion characteristics while maintaining low optical losses, thereby enabling a wide bandwidth exceeding 2.5 THz with a polarization ratio surpassing 17 dB. Moreover, the operational frequency can be adjusted across a 3.4 THz range by altering the incident angle of the incident waves. This adaptable design, characterized by its polarization versatility, can be customized to suit practical applications within wireless communication, radar systems, optical switching, and imaging technologies., Comment: Accepted by Applied Physics Letters

Tunable optical filter at the chip scale plays a crucial role in fulfilling the need for the reconfigurability in channel routing, optical switching, and wavelength division multiplexing systems. In this letter, we propose a tunable double notch filter on thin-film lithium niobate using dual micro-ring architecture. This unique integrated filter is essential for complex photonic integrated circuits, along with multiple channels and various frequency spacing. With only one loaded voltage, the device demonstrates a wide frequency spacing tunability from 16.1 GHz to 89.9 GHz by reversely tunning the resonances of the two micro-rings while the center wavelength between the two resonances remains unaltered. Moreover, by utilizing the pronounced electro-optic properties of lithium niobate, associated with the tight light confinement nanophotonic waveguides, the device demonstrates a spacing tunability of 0.82 GHz/V and a contrast of 10~16 dB. In addition, the device has an ultracompact footprint of 0.0248 mm2., Comment: accepted by Optics Letters

We develop a new method called affine FR for recovering Slater's condition for semidefinite programming (SDP) relaxations of combinatorial optimization (CO) problems. Affine FR is a user-friendly method, as it is fully automatic and only requires a description of the problem. We provide a rigorous analysis of differences between affine FR and the existing methods. We also present numerical results to demonstrate the effectiveness of affine FR in reducing the size of SDP relaxations for CO problems.

The singularity degree plays a crucial role in understanding linear and semidefinite programming, providing a theoretical framework for analyzing these problems. It is defined as the minimum number of facial reduction (FR) steps needed to reach strict feasibility for a convex set. On the other hand, the maximum singularity degree (MSD) is the maximum number of steps required. Recent progress in the applications of MSD has motivated us to explore its fundamental properties in this paper. For semidefinite programming, we establish a necessary condition for an FR sequence to be the longest. Additionally, we propose an upper bound for MSD, which can be computed more easily. By leveraging these findings, we prove that computing MSD is NP-hard. This complexity result complements the existing algorithms for computing the singularity degree found in the literature. For linear programming, we provide a characterization for the longest FR sequences, which also serves as a polynomial-time algorithm for constructing such a sequence. In addition, we introduce two operations that ensure the longest FR sequences remain the longest. Lastly, we prove that MSD is equivalent to a novel parameter called the implicit problem singularity.

The computational demands of modern AI have spurred interest in optical neural networks (ONNs) which offer the potential benefits of increased speed and lower power consumption. However, current ONNs face various challenges,most significantly a limited calculation precision (typically around 4 bits) and the requirement for high-resolution signal format converters (digital-to-analogue conversions (DACs) and analogue-to-digital conversions (ADCs)). These challenges are inherent to their analog computing nature and pose significant obstacles in practical implementation. Here, we propose a digital-analog hybrid optical computing architecture for ONNs, which utilizes digital optical inputs in the form of binary words. By introducing the logic levels and decisions based on thresholding, the calculation precision can be significantly enhanced. The DACs for input data can be removed and the resolution of the ADCs can be greatly reduced. This can increase the operating speed at a high calculation precision and facilitate the compatibility with microelectronics. To validate our approach, we have fabricated a proof-of-concept photonic chip and built up a hybrid optical processor (HOP) system for neural network applications. We have demonstrated an unprecedented 16-bit calculation precision for high-definition image processing, with a pixel error rate (PER) as low as $1.8\times10^{-3}$ at an signal-to-noise ratio (SNR) of 18.2 dB. We have also implemented a convolutional neural network for handwritten digit recognition that shows the same accuracy as the one achieved by a desktop computer. The concept of the digital-analog hybrid optical computing architecture offers a methodology that could potentially be applied to various ONN implementations and may intrigue new research into efficient and accurate domain-specific optical computing architectures for neural networks.

This paper focuses on the analysis of the application effectiveness of the integration of deep learning and computer vision technologies. Deep learning achieves a historic breakthrough by constructing hierarchical neural networks, enabling end-to-end feature learning and semantic understanding of images. The successful experiences in the field of computer vision provide strong support for training deep learning algorithms. The tight integration of these two fields has given rise to a new generation of advanced computer vision systems, significantly surpassing traditional methods in tasks such as machine vision image classification and object detection. In this paper, typical image classification cases are combined to analyze the superior performance of deep neural network models while also pointing out their limitations in generalization and interpretability, proposing directions for future improvements. Overall, the efficient integration and development trend of deep learning with massive visual data will continue to drive technological breakthroughs and application expansion in the field of computer vision, making it possible to build truly intelligent machine vision systems. This deepening fusion paradigm will powerfully promote unprecedented tasks and functions in computer vision, providing stronger development momentum for related disciplines and industries.

High-fidelity and efficient audio-driven talking head generation has been a key research topic in computer graphics and computer vision. In this work, we study vector image based audio-driven talking head generation. Compared with directly animating the raster image that most widely used in existing works, vector image enjoys its excellent scalability being used for many applications. There are two main challenges for vector image based talking head generation: the high-quality vector image reconstruction w.r.t. the source portrait image and the vivid animation w.r.t. the audio signal. To address these, we propose a novel scalable vector graphic reconstruction and animation method, dubbed VectorTalker. Specifically, for the highfidelity reconstruction, VectorTalker hierarchically reconstructs the vector image in a coarse-to-fine manner. For the vivid audio-driven facial animation, we propose to use facial landmarks as intermediate motion representation and propose an efficient landmark-driven vector image deformation module. Our approach can handle various styles of portrait images within a unified framework, including Japanese manga, cartoon, and photorealistic images. We conduct extensive quantitative and qualitative evaluations and the experimental results demonstrate the superiority of VectorTalker in both vector graphic reconstruction and audio-driven animation.

Intelligent reflecting surface (IRS)-assisted unmanned aerial vehicle (UAV) communications are expected to alleviate the load of ground base stations in a cost-effective way. Existing studies mainly focus on the deployment and resource allocation of a single IRS instead of multiple IRSs, whereas it is extremely challenging for joint multi-IRS multi-user association in UAV communications with constrained reflecting resources and dynamic scenarios. To address the aforementioned challenges, we propose a new optimization algorithm for joint IRS-user association, trajectory optimization of UAVs, successive interference cancellation (SIC) decoding order scheduling and power allocation to maximize system energy efficiency. We first propose an inverse soft-Q learning-based algorithm to optimize multi-IRS multi-user association. Then, SCA and Dinkelbach-based algorithm are leveraged to optimize UAV trajectory followed by the optimization of SIC decoding order scheduling and power allocation. Finally, theoretical analysis and performance results show significant advantages of the designed algorithm in convergence rate and energy efficiency.

Compared with conventional grating-based spectrometers, reconstructive spectrometers based on spectrally engineered filtering have the advantage of miniaturization because of the less demand for dispersive optics and free propagation space. However, available reconstructive spectrometers fail to balance the performance on operational bandwidth, spectral diversity and angular stability. In this work, we proposed a compact silicon metasurfaces based spectrometer/camera. After angle integration, the spectral response of the system is robust to angle/aperture within a wide working bandwidth from 400nm to 800nm. It is experimentally demonstrated that the proposed method could maintain the spectral consistency from F/1.8 to F/4 (The corresponding angle of incident light ranges from 7{\deg} to 16{\deg}) and the incident hyperspectral signal could be accurately reconstructed with a fidelity exceeding 99%. Additionally, a spectral imaging system with 400x400 pixels is also established in this work. The accurate reconstructed hyperspectral image indicates that the proposed aperture-robust spectrometer has the potential to be extended as a high-resolution broadband hyperspectral camera.

Reward-free data is abundant and contains rich prior knowledge of human behaviors, but it is not well exploited by offline reinforcement learning (RL) algorithms. In this paper, we propose UBER, an unsupervised approach to extract useful behaviors from offline reward-free datasets via diversified rewards. UBER assigns different pseudo-rewards sampled from a given prior distribution to different agents to extract a diverse set of behaviors, and reuse them as candidate policies to facilitate the learning of new tasks. Perhaps surprisingly, we show that rewards generated from random neural networks are sufficient to extract diverse and useful behaviors, some even close to expert ones. We provide both empirical and theoretical evidence to justify the use of random priors for the reward function. Experiments on multiple benchmarks showcase UBER's ability to learn effective and diverse behavior sets that enhance sample efficiency for online RL, outperforming existing baselines. By reducing reliance on human supervision, UBER broadens the applicability of RL to real-world scenarios with abundant reward-free data., Comment: Thirty-seventh Conference on Neural Information Processing Systems

van der Waals (vdW) layered materials exhibit fruitful novel physical properties. The energy band of such materials depends strongly on their structures and a tremendous variation in their physical properties can be deduced from a tiny change in inter-layer spacing, twist angle, or in-plane strain. In this work, a kind of vdW layered material of spiral antimonene is constructed, and the strain effects in the material are studied. The spiral antimonene is grown on a germanium (Ge) substrate and is induced by a helical dislocation penetrating through few-atomic-layers of antimonene (\b{eta}-phase). The as-grown spiral is intrinsically strained and the lattice distortion is found to be pinned around the dislocation. Both spontaneous inter-layer twist and in-plane anisotropic strain are observed in scanning tunneling microscope (STM) measurements. The strain in the spiral antimonene can be significantly modified by STM tip interaction, leading to a variation in the surface electronic density of states (DOS) and a large modification in the work function of up to a few hundreds of milli-electron-volts (meV). Those strain effects are expected to have potential applications in building up novel piezoelectric devices., Comment: 11 pages, 4 figures, Supporting Information

Neural Radiance Fields (NeRFs) are gaining significant interest for online active object reconstruction due to their exceptional memory efficiency and requirement for only posed RGB inputs. Previous NeRF-based view planning methods exhibit computational inefficiency since they rely on an iterative paradigm, consisting of (1) retraining the NeRF when new images arrive; and (2) planning a path to the next best view only. To address these limitations, we propose a non-iterative pipeline based on the Prediction of the Required number of Views (PRV). The key idea behind our approach is that the required number of views to reconstruct an object depends on its complexity. Therefore, we design a deep neural network, named PRVNet, to predict the required number of views, allowing us to tailor the data acquisition based on the object complexity and plan a globally shortest path. To train our PRVNet, we generate supervision labels using the ShapeNet dataset. Simulated experiments show that our PRV-based view planning method outperforms baselines, achieving good reconstruction quality while significantly reducing movement cost and planning time. We further justify the generalization ability of our approach in a real-world experiment., Comment: Sicong Pan and Liren Jin have equal contribution. Publication to appear in IEEE International Conference on Robotics and Automation (ICRA), 2024

Active object reconstruction using autonomous robots is gaining great interest. A primary goal in this task is to maximize the information of the object to be reconstructed, given limited on-board resources. Previous view planning methods exhibit inefficiency since they rely on an iterative paradigm based on explicit representations, consisting of (1) planning a path to the next-best view only; and (2) requiring a considerable number of less-gain views in terms of surface coverage. To address these limitations, we propose to integrate implicit representations into the One-Shot View Planning (OSVP). The key idea behind our approach is to use implicit representations to obtain the small missing surface areas instead of observing them with extra views. Therefore, we design a deep neural network, named OSVP, to directly predict a set of views given a dense point cloud refined from an initial sparse observation. To train our OSVP network, we generate supervision labels using dense point clouds refined by implicit representations and set covering optimization problems. Simulated experiments show that our method achieves sufficient reconstruction quality, outperforming several baselines under limited view and movement budgets. We further demonstrate the applicability of our approach in a real-world object reconstruction scenario., Comment: Hao Hu and Sicong Pan have equal contribution. Publication to appear in IEEE International Conference on Robotics and Automation (ICRA), 2024

Large language models (LLMs) demonstrate impressive reasoning abilities, but translating reasoning into actions in the real world remains challenging. In particular, it remains unclear how to complete a given task provably within a minimum number of interactions with the external environment, e.g., through an internal mechanism of reasoning. To this end, we propose a principled framework with provable regret guarantees to orchestrate reasoning and acting, which we call "reason for future, act for now" (\texttt{RAFA}). Specifically, we design a prompt template for reasoning that learns from the memory buffer and plans a future trajectory over a long horizon ("reason for future"). At each step, the LLM agent takes the initial action of the planned trajectory ("act for now"), stores the collected feedback in the memory buffer, and reinvokes the reasoning routine to replan the future trajectory from the new state. The key idea is to cast reasoning in LLMs as learning and planning in Bayesian adaptive Markov decision processes (MDPs). Correspondingly, we prompt LLMs to form an updated posterior of the unknown environment from the memory buffer (learning) and generate an optimal trajectory for multiple future steps that maximizes a value function (planning). The learning and planning subroutines are performed in an "in-context" manner to emulate the actor-critic update for MDPs. Our theoretical analysis proves that the novel combination of long-term reasoning and short-term acting achieves a $\sqrt{T}$ regret. Here, $T$ denotes the number of online interactions. In particular, the regret bound highlights an intriguing interplay between the prior knowledge obtained through pretraining and the uncertainty reduction achieved by reasoning and acting. Our empirical validation shows that it outperforms various existing frameworks and achieves nearly perfect scores on a few benchmarks., Comment: Accepted by International Conference on Machine Learning (ICML) 2024

The Sweeny algorithm for the $Q$-state random-cluster model in two dimensions is shown to exhibit a rich mixture of critical dynamical scaling behaviors. As $Q$ decreases, the so-called critical speeding-up for non-local quantities becomes more and more pronounced. However, for some quantity of specific local pattern -- e.g., the number of half faces on the square lattice, we observe that, as $Q \to 0$, the integrated autocorrelation time $\tau$ diverges as $Q^{-\zeta}$, with $\zeta \simeq 1/2$, leading to the non-ergodicity of the Sweeny method for $Q \to 0$. Such $Q$-dependent critical slowing-down, attributed to the peculiar form of the critical bond weight $v=\sqrt{Q}$, can be eliminated by a combination of the Sweeny and the Kawasaki algorithm. Moreover, by classifying the occupied bonds into bridge bonds and backbone bonds, and the empty bonds into internal-perimeter bonds and external-perimeter bonds, one can formulate an improved version of the Sweeny-Kawasaki method such that the autocorrelation time for any quantity is of order $O(1)$., Comment: 10 pages, 8 figures, accepted for publication in Physical Review E

Many successful machine learning methods have been developed for microwave inversion problems. However, so far, their inversion has been performed only at the specifically trained frequencies. To make the machine-learning-based inversion method more generalizability for realistic engineering applications, this work proposes a residual fully convolutional network (Res-FCN) to perform microwave inversion of high contrast scatterers at an arbitrary frequency within a wide frequency band. The proposed Res-FCN combines the advantages of the Res-Net and the fully convolutional network (FCN). Res-FCN consists of an encoder and a decoder: the encoder is employed to extract high-dimensional features from the measured scattered field through the residual frameworks, while the decoder is employed to map from the high-dimensional features extracted by the encoder to the electrical parameter distribution in the inversion region by the up-sample layer and the residual frameworks. Five numerical examples verify that the proposed Res-FCN can achieve good performance in the 2-D microwave inversion problem for high contrast scatterers with anti-noise ability at an arbitrary frequency point within a wide frequency band.

Offline goal-conditioned RL (GCRL) offers a way to train general-purpose agents from fully offline datasets. In addition to being conservative within the dataset, the generalization ability to achieve unseen goals is another fundamental challenge for offline GCRL. However, to the best of our knowledge, this problem has not been well studied yet. In this paper, we study out-of-distribution (OOD) generalization of offline GCRL both theoretically and empirically to identify factors that are important. In a number of experiments, we observe that weighted imitation learning enjoys better generalization than pessimism-based offline RL method. Based on this insight, we derive a theory for OOD generalization, which characterizes several important design choices. We then propose a new offline GCRL method, Generalizable Offline goAl-condiTioned RL (GOAT), by combining the findings from our theoretical and empirical studies. On a new benchmark containing 9 independent identically distributed (IID) tasks and 17 OOD tasks, GOAT outperforms current state-of-the-art methods by a large margin., Comment: Accepted by International Conference on Machine Learning (ICML), 2023

In online reinforcement learning (online RL), balancing exploration and exploitation is crucial for finding an optimal policy in a sample-efficient way. To achieve this, existing sample-efficient online RL algorithms typically consist of three components: estimation, planning, and exploration. However, in order to cope with general function approximators, most of them involve impractical algorithmic components to incentivize exploration, such as optimization within data-dependent level-sets or complicated sampling procedures. To address this challenge, we propose an easy-to-implement RL framework called \textit{Maximize to Explore} (\texttt{MEX}), which only needs to optimize \emph{unconstrainedly} a single objective that integrates the estimation and planning components while balancing exploration and exploitation automatically. Theoretically, we prove that \texttt{MEX} achieves a sublinear regret with general function approximations for Markov decision processes (MDP) and is further extendable to two-player zero-sum Markov games (MG). Meanwhile, we adapt deep RL baselines to design practical versions of \texttt{MEX}, in both model-free and model-based manners, which can outperform baselines by a stable margin in various MuJoCo environments with sparse rewards. Compared with existing sample-efficient online RL algorithms with general function approximations, \texttt{MEX} achieves similar sample efficiency while enjoying a lower computational cost and is more compatible with modern deep RL methods.

Existing view planning systems either adopt an iterative paradigm using next-best views (NBV) or a one-shot pipeline relying on the set-covering view-planning (SCVP) network. However, neither of these methods can concurrently guarantee both high-quality and high-efficiency reconstruction of 3D unknown objects. To tackle this challenge, we introduce a crucial hypothesis: with the availability of more information about the unknown object, the prediction quality of the SCVP network improves. There are two ways to provide extra information: (1) leveraging perception data obtained from NBVs, and (2) training on an expanded dataset of multiview inputs. In this work, we introduce a novel combined pipeline that incorporates a single NBV before activating the proposed multiview-activated (MA-)SCVP network. The MA-SCVP is trained on a multiview dataset generated by our long-tail sampling method, which addresses the issue of unbalanced multiview inputs and enhances the network performance. Extensive simulated experiments substantiate that our system demonstrates a significant surface coverage increase and a substantial 45% reduction in movement cost compared to state-of-the-art systems. Real-world experiments justify the capability of our system for generalization and deployment., Comment: Revised and Resubmitted to IEEE Transaction on Robotics

Effective-medium theory pertains to the theoretical modelling of homogenization, which aims to replace an inhomogeneous structure of subwavelength-scale constituents with a homogeneous effective medium. The effective-medium theory is fundamental to various realms, including electromagnetics and material science, since it can largely decrease the complexity in the exploration of light-matter interactions by providing simple acceptable approximation. Generally, the effective-medium theory is thought to be applicable to any all-dielectric system with deep-subwavelength constituents, under the condition that the effective medium does not have a critical angle, at which the total internal reflection occurs. Here we reveal a fundamental breakdown of the effective-medium theory that can be applied in very general conditions: showing it for deep-subwavelength all-dielectric multilayers even without critical angle. Our finding relies on an exotic photonic spin Hall effect, which is shown to be ultra-sensitive to the stacking order of deep-subwavelength dielectric layers, since the spin-orbit interaction of light is dependent on slight phase accumulations during the wave propagation. Our results indicate that the photonic spin Hall effect could provide a promising and powerful tool for measuring structural defects for all-dielectric systems even in the extreme nanometer scale., Comment: 20 pages, 5 figures