Search

ATARI is a suite of video games used by reinforcement learning (RL) researchers to test the effectiveness of the learning algorithm. Receiving only the raw pixels and the game score, the agent learns to develop sophisticated strategies, even to the comparable level of a professional human games tester. Ideally, we also want an agent requiring very few interactions with the environment. Previous competitive model-free algorithms for the task use the valued-based Rainbow algorithm without any policy head. In this paper, we change it by proposing a practical discrete variant of the soft actor-critic (SAC) algorithm. The new variant enables off-policy learning using policy heads for discrete domains. By incorporating it into the advanced Rainbow variant, i.e., the ``bigger, better, faster'' (BBF), the resulting SAC-BBF improves the previous state-of-the-art interquartile mean (IQM) from 1.045 to 1.088, and it achieves these results using only replay ratio (RR) 2. By using lower RR 2, the training time of SAC-BBF is strictly one-third of the time required for BBF to achieve an IQM of 1.045 using RR 8. As a value of IQM greater than one indicates super-human performance, SAC-BBF is also the only model-free algorithm with a super-human level using only RR 2. The code is publicly available on GitHub at https://github.com/lezhang-thu/bigger-better-faster-SAC., Comment: Chinese Conference on Pattern Recognition and Computer Vision (PRCV) 2024. GitHub Repo https://github.com/lezhang-thu/bigger-better-faster-SAC

Recently, both global and local classical randomness-assisted projective measurement protocols have been employed to share Bell nonlocality of an entangled state among multiple sequential parties. Unlike Bell nonlocality, Einstein-Podolsky-Rosen (EPR) steering exhibits distinct asymmetric characteristics and serves as the necessary quantum resource for one-sided device-independent quantum information tasks. In this work, we propose a projective measurement protocol and investigate the shareability of EPR steering with steering radius criterion theoretically and experimentally. Our results reveal that arbitrarily many independent parties can share one-way steerability using projective measurements, even when no shared randomness is available. Furthermore, by leveraging only local randomness, asymmetric two-way steerability can also be shared. Our work not only deepens the understanding of the role of projective measurements in sharing quantum correlations but also opens up a new avenue for reutilizing asymmetric quantum correlations., Comment: 13 pages, 6 figures

There have been many image denoisers using deep neural networks, which outperform conventional model-based methods by large margins. Recently, self-supervised methods have attracted attention because constructing a large real noise dataset for supervised training is an enormous burden. The most representative self-supervised denoisers are based on blind-spot networks, which exclude the receptive field's center pixel. However, excluding any input pixel is abandoning some information, especially when the input pixel at the corresponding output position is excluded. In addition, a standard blind-spot network fails to reduce real camera noise due to the pixel-wise correlation of noise, though it successfully removes independently distributed synthetic noise. Hence, to realize a more practical denoiser, we propose a novel self-supervised training framework that can remove real noise. For this, we derive the theoretic upper bound of a supervised loss where the network is guided by the downsampled blinded output. Also, we design a conditional blind-spot network (C-BSN), which selectively controls the blindness of the network to use the center pixel information. Furthermore, we exploit a random subsampler to decorrelate noise spatially, making the C-BSN free of visual artifacts that were often seen in downsample-based methods. Extensive experiments show that the proposed C-BSN achieves state-of-the-art performance on real-world datasets as a self-supervised denoiser and shows qualitatively pleasing results without any post-processing or refinement., Comment: Accepted to ICCV 2023

Quantum steering is commonly shared among multiple observers by utilizing unsharp measurements. However, their usage is limited to local measurements and is not suitable for nonlocal-measurement-based cases. Here, we present a novel approach in this study, suggesting a highly efficient technique to construct optimal nonlocal measurements by utilizing quantum ellipsoids to share quantum steering. This technique is suitable for any bipartite state and offers benefits even in scenarios with a high number of measurement settings. Using the Greenberger-Horne-Zeilinger state as an illustration, we show that employing unsharp nonlocal product measurements can activate the phenomenon of steering sharing in contrast to using local measurements. Moreover, our findings demonstrate that nonlocal measurements with unequal strength possess a greater activation capability compared to those with equal strength. Our activation method differs from previous ones as it eliminates the need to copy the shared states or diminish other quantum correlations, thus making it convenient for practical experimentation and conservation of resources., Comment: 9 pages, 4 figures

As a way of saving quantum resources, recycling of Bell nonlocality has been experimentally studied, but restricted to sequential unsharp measurements. However, it has been theoretically shown recently that projective measurements are sufficient for recycling nonlocality [Phys. Rev. Lett. \textbf{129}, 230402 (2022)]. Here, we go beyond unsharp measurement scenarios and experimentally demonstrate the recycling of nonlocal resources with projective measurements. By verifying the violation of Clauser-Horne-Shimony-Holt (CHSH) inequality, we find that three independent parties can recycle the Bell nonlocality of a two-qubit state, whether it is maximally or partially entangled. Furthermore, in the double violation region, the optimal trade-off for partially entangled states can be 11 standard deviations better than that for maximally entangled states. Our results experimentally eliminate the common misconception that projective measurements are incompatible with the recycling of quantum correlations. In addition, our nonlocality recycling setup does not require entanglement assistance, which is much more experimentally friendly, thus paving the way for the reuse of other kinds of quantum correlations., Comment: 8 pages, 5 figures

High-precision surface roughness estimation plays an important role in many applications. However, the classical estimating methods are limited by shot noise and only can achieve the precision of 0.1 nm with white light interferometer. Here, we propose two weak measurement schemes to estimate surface roughness through spectrum analysis and intensity analysis. The estimating precision with spectrum analysis is about $10 ^{-5}$ nm by using a currently available spectrometer with the resolution of $\Delta \lambda= 0.04$ pm and the corresponding sensitivity is better than 0.1 THz/nm. And the precision and sensitivity of the light intensity analysis scheme achieve as high as 0.07 nm and 1/nm, respectively. By introducing a modulated phase, we show that the sensitivity and precision achieved in our schemes can be effectively retained in a wider dynamic range. We further provide the experimental design of the surface profiler based on our schemes. It simultaneously meets the requirements of high precision, high sensitivity, and wide measurement range, making it to be a promising practical tool., Comment: 7 pages, 6 figures

The sequential quantum random access code (QRAC) allows two or more decoders to obtain a desired message with higher success probability than the best classical bounds by appropriately modulating the measurement sharpness. Here, we propose an entanglement-assisted sequential QRAC protocol which can enable device-independent tasks. By relaxing the equal sharpness and mutually unbiased measurement limits, we widen the sharpness modulation region from a one-dimensional interval to a two-dimensional triangle. Then, we demonstrate our scheme experimentally and get more than 27 standard deviations above the classical bound even when both decoders perform approximately projective measurements. We use the observed success probability to quantify the connection among sequential QRAC, measurement sharpness, measurement biasedness, and measurement incompatibility. Finally, we show that our protocol can be applied to sequential device-independent randomness expansion and our measurement strategy can enhance the success probability of decoding the entire input string. Our results may promote a deeper understanding of the relationship among quantum correlation, quantum measurement, and quantum information processing., Comment: 10 pages, 4 figures, 8 tables

We continue the study of distance sensitivity oracles (DSOs). Given a directed graph $G$ with $n$ vertices and edge weights in $\{1, 2, \dots, M\}$, we want to build a data structure such that given any source vertex $u$, any target vertex $v$, and any failure $f$ (which is either a vertex or an edge), it outputs the length of the shortest path from $u$ to $v$ not going through $f$. Our main result is a DSO with preprocessing time $O(n^{2.5794}M)$ and constant query time. Previously, the best preprocessing time of DSOs for directed graphs is $O(n^{2.7233}M)$, and even in the easier case of undirected graphs, the best preprocessing time is $O(n^{2.6865}M)$ [Ren, ESA 2020]. One drawback of our DSOs, though, is that it only supports distance queries but not path queries. Our main technical ingredient is an algorithm that computes the inverse of a degree-$d$ polynomial matrix (i.e. a matrix whose entries are degree-$d$ univariate polynomials) modulo $x^r$. The algorithm is adapted from [Zhou, Labahn, and Storjohann, Journal of Complexity, 2015], and we replace some of its intermediate steps with faster rectangular matrix multiplication algorithms. We also show how to compute unique shortest paths in a directed graph with edge weights in $\{1, 2, \dots, M\}$, in $O(n^{2.5286}M)$ time. This algorithm is crucial in the preprocessing algorithm of our DSO. Our solution improves the $O(n^{2.6865}M)$ time bound in [Ren, ESA 2020], and matches the current best time bound for computing all-pairs shortest paths., Comment: accepted to ICALP 2021

In this paper, we propose an efficient and accurate method for autonomous surface vehicles to generate a smooth and collision-free trajectory considering its dynamics constraints. We decouple the trajectory planning problem as a front-end feasible path searching and a back-end kinodynamic trajectory optimization. Firstly, we model the type of two-thrusts under-actuated surface vessel. Then we adopt a sampling-based path searching to find an asymptotic optimal path through the obstacle-surrounding environment and extract several waypoints from it. We apply a numerical optimization method in the back-end to generate the trajectory. From the perspective of security in the field voyage, we propose the sailing corridor method to guarantee the trajectory away from obstacles. Moreover, considering limited fuel ASV carrying, we design a numerical objective function which can optimize a fuel-saving trajectory. Finally, we validate and compare the proposed method in simulation environments and the results fit our expected trajectory.

Real-noise denoising is a challenging task because the statistics of real-noise do not follow the normal distribution, and they are also spatially and temporally changing. In order to cope with various and complex real-noise, we propose a well-generalized denoising architecture and a transfer learning scheme. Specifically, we adopt an adaptive instance normalization to build a denoiser, which can regularize the feature map and prevent the network from overfitting to the training set. We also introduce a transfer learning scheme that transfers knowledge learned from synthetic-noise data to the real-noise denoiser. From the proposed transfer learning, the synthetic-noise denoiser can learn general features from various synthetic-noise data, and the real-noise denoiser can learn the real-noise characteristics from real data. From the experiments, we find that the proposed denoising method has great generalization ability, such that our network trained with synthetic-noise achieves the best performance for Darmstadt Noise Dataset (DND) among the methods from published papers. We can also see that the proposed transfer learning scheme robustly works for real-noise images through the learning with a very small number of labeled data., Comment: CVPR accepted paper. The paper will be updated according to reviewers' comments

Given an undirected graph $G=(V,E)$ of $n$ vertices and $m$ edges with weights in $[1,W]$, we construct vertex sensitive distance oracles (VSDO), which are data structures that preprocess the graph, and answer the following kind of queries: Given a source vertex $u$, a target vertex $v$, and a batch of $d$ failed vertices $D$, output (an approximation of) the distance between $u$ and $v$ in $G-D$ (that is, the graph $G$ with vertices in $D$ removed). An oracle has stretch $\alpha$ if it always holds that $\delta_{G-D}(u,v)\le\tilde{\delta}(u,v)\le\alpha\cdot\delta_{G-D}(u,v)$, where $\delta_{G-D}(u,v)$ is the actual distance between $u$ and $v$ in $G-D$, and $\tilde{\delta}(u,v)$ is the distance reported by the oracle. In this paper we construct efficient VSDOs for any number $d$ of failures. For any constant $c\geq 1$, we propose two oracles: $\bullet$ The first oracle has size $n^{2+1/c}(\log n/\epsilon)^{O(d)}\cdot \log W$, answers a query in ${\rm poly}(\log n,d^c,\log\log W,\epsilon^{-1})$ time, and has stretch $1+\epsilon$, for any constant $\epsilon>0$. $\bullet$ The second oracle has size $n^{2+1/c}{\rm poly}(\log (nW),d)$, answers a query in ${\rm poly}(\log n,d^c,\log\log W)$ time, and has stretch ${\rm poly}(\log n,d)$. Both of these oracles can be preprocessed in time polynomial in their space complexity. These results are the first approximate distance oracles of poly-logarithmic query time for any constant number of vertex failures in general undirected graphs. Previously there are $(1+\epsilon)$-approximate $d$-edge sensitive distance oracles [Chechik et al. 2017] answering distance queries when $d$ edges fail, which have size $O(n^2(\log n/\epsilon)^d\cdot d\log W)$ and query time ${\rm poly}(\log n, d, \log\log W)$., Comment: SODA'21

A roundtrip spanner of a directed graph $G$ is a subgraph of $G$ preserving roundtrip distances approximately for all pairs of vertices. Despite extensive research, there is still a small stretch gap between roundtrip spanners in directed graphs and undirected graphs. For a directed graph with real edge weights in $[1,W]$, we first propose a new deterministic algorithm that constructs a roundtrip spanner with $(2k-1)$ stretch and $O(k n^{1+1/k}\log (nW))$ edges for every integer $k> 1$, then remove the dependence of size on $W$ to give a roundtrip spanner with $(2k-1)$ stretch and $O(k n^{1+1/k}\log n)$ edges. While keeping the edge size small, our result improves the previous $2k+\epsilon$ stretch roundtrip spanners in directed graphs [Roditty, Thorup, Zwick'02; Zhu, Lam'18], and almost matches the undirected $(2k-1)$-spanner with $O(n^{1+1/k})$ edges [Alth\"ofer et al. '93] when $k$ is a constant, which is optimal under Erd\"os conjecture., Comment: 12 pages

Recently, many convolutional neural networks for single image super-resolution (SISR) have been proposed, which focus on reconstructing the high-resolution images in terms of objective distortion measures. However, the networks trained with objective loss functions generally fail to reconstruct the realistic fine textures and details that are essential for better perceptual quality. Recovering the realistic details remains a challenging problem, and only a few works have been proposed which aim at increasing the perceptual quality by generating enhanced textures. However, the generated fake details often make undesirable artifacts and the overall image looks somewhat unnatural. Therefore, in this paper, we present a new approach to reconstructing realistic super-resolved images with high perceptual quality, while maintaining the naturalness of the result. In particular, we focus on the domain prior properties of SISR problem. Specifically, we define the naturalness prior in the low-level domain and constrain the output image in the natural manifold, which eventually generates more natural and realistic images. Our results show better naturalness compared to the recent super-resolution algorithms including perception-oriented ones., Comment: Presented in CVPR 2019