arXiv subjects

Haoran Wang

Publications and source records attributed to Haoran Wang.

229 records · Page 13Linked to original sources

arXiv2020

UDC 2020 Challenge on Image Restoration of Under-Display Camera: Methods and Results

This paper is the report of the first Under-Display Camera (UDC) image restoration challenge in conjunction with the RLQ workshop at ECCV 2020. The challenge is based on a newly-collected database of Under-Display Camera. The challenge tracks correspond to two types of display: a 4k Transparent OLED (T-OLED) and a phone Pentile OLED (P-OLED). Along with about 150 teams registered the challenge, eight and nine teams submitted the results during the testing phase for each track. The results in the paper are state-of-the-art restoration performance of Under-Display Camera Restoration. Datasets and paper are available at https://yzhouas.github.io/projects/UDC/udc.html.

eess.IV↗

arXiv2020

Classes Matter: A Fine-grained Adversarial Approach to Cross-domain Semantic Segmentation

Despite great progress in supervised semantic segmentation,a large performance drop is usually observed when deploying the model in the wild. Domain adaptation methods tackle the issue by aligning the source domain and the target domain. However, most existing methods attempt to perform the alignment from a holistic view, ignoring the underlying class-level data structure in the target domain. To fully exploit the supervision in the source domain, we propose a fine-grained adversarial learning strategy for class-level feature alignment while preserving the internal structure of semantics across domains. We adopt a fine-grained domain discriminator that not only plays as a domain distinguisher, but also differentiates domains at class level. The traditional binary domain labels are also generalized to domain encodings as the supervision signal to guide the fine-grained feature alignment. An analysis with Class Center Distance (CCD) validates that our fine-grained adversarial strategy achieves better class-level alignment compared to other state-of-the-art methods. Our method is easy to implement and its effectiveness is evaluated on three classical domain adaptation tasks, i.e., GTA5 to Cityscapes, SYNTHIA to Cityscapes and Cityscapes to Cross-City. Large performance gains show that our method outperforms other global feature alignment based and class-wise alignment based counterparts. The code is publicly available at https://github.com/JDAI-CV/FADA.

cs.CV↗

arXiv2020

Intrinsic and Extrinsic Approximation of Koopman Operators over Manifolds

This paper derives rates of convergence of certain approximations of the Koopman operators that are associated with discrete, deterministic, continuous semiflows on a complete metric space $(X,d_X)$. Approximations are constructed in terms of reproducing kernel bases that are centered at samples taken along the system trajectory. It is proven that when the samples are dense in a certain type of smooth manifold $M\subseteq X$, the derived rates of convergence depend on the fill distance of samples along the trajectory in that manifold. Error bounds for projection-based and data-dependent approximations of the Koopman operator are derived in the paper. A discussion of how these bounds are realized in intrinsic and extrinsic approximation methods is given. Finally, a numerical example that illustrates qualitatively the convergence guarantees derived in the paper is given.

math.DS↗

arXiv2019

Large scale continuous-time mean-variance portfolio allocation via reinforcement learning

We propose to solve large scale Markowitz mean-variance (MV) portfolio allocation problem using reinforcement learning (RL). By adopting the recently developed continuous-time exploratory control framework, we formulate the exploratory MV problem in high dimensions. We further show the optimality of a multivariate Gaussian feedback policy, with time-decaying variance, in trading off exploration and exploitation. Based on a provable policy improvement theorem, we devise a scalable and data-efficient RL algorithm and conduct large scale empirical tests using data from the S&P 500 stocks. We found that our method consistently achieves over 10% annualized returns and it outperforms econometric methods and the deep RL method by large margins, for both long and medium terms of investment with monthly and daily trading.

q-fin.PM↗

arXiv2019

Continuous-Time Mean-Variance Portfolio Selection: A Reinforcement Learning Framework

We approach the continuous-time mean-variance (MV) portfolio selection with reinforcement learning (RL). The problem is to achieve the best tradeoff between exploration and exploitation, and is formulated as an entropy-regularized, relaxed stochastic control problem. We prove that the optimal feedback policy for this problem must be Gaussian, with time-decaying variance. We then establish connections between the entropy-regularized MV and the classical MV, including the solvability equivalence and the convergence as exploration weighting parameter decays to zero. Finally, we prove a policy improvement theorem, based on which we devise an implementable RL algorithm. We find that our algorithm outperforms both an adaptive control based method and a deep neural networks based algorithm by a large margin in our simulations.

q-fin.PM↗

arXiv2019

Exploration versus exploitation in reinforcement learning: a stochastic control approach

We consider reinforcement learning (RL) in continuous time and study the problem of achieving the best trade-off between exploration of a black box environment and exploitation of current knowledge. We propose an entropy-regularized reward function involving the differential entropy of the distributions of actions, and motivate and devise an exploratory formulation for the feature dynamics that captures repetitive learning under exploration. The resulting optimization problem is a revitalization of the classical relaxed stochastic control. We carry out a complete analysis of the problem in the linear--quadratic (LQ) setting and deduce that the optimal feedback control distribution for balancing exploitation and exploration is Gaussian. This in turn interprets and justifies the widely adopted Gaussian exploration in RL, beyond its simplicity for sampling. Moreover, the exploitation and exploration are captured, respectively and mutual-exclusively, by the mean and variance of the Gaussian distribution. We also find that a more random environment contains more learning opportunities in the sense that less exploration is needed. We characterize the cost of exploration, which, for the LQ case, is shown to be proportional to the entropy regularization weight and inversely proportional to the discount rate. Finally, as the weight of exploration decays to zero, we prove the convergence of the solution of the entropy-regularized LQ problem to the one of the classical LQ problem.

math.OC↗

arXiv2018

Parameter-Free Spatial Attention Network for Person Re-Identification

Global average pooling (GAP) allows to localize discriminative information for recognition [40]. While GAP helps the convolution neural network to attend to the most discriminative features of an object, it may suffer if that information is missing e.g. due to camera viewpoint changes. To circumvent this issue, we argue that it is advantageous to attend to the global configuration of the object by modeling spatial relations among high-level features. We propose a novel architecture for Person Re-Identification, based on a novel parameter-free spatial attention layer introducing spatial relations among the feature map activations back to the model. Our spatial attention layer consistently improves the performance over the model without it. Results on four benchmarks demonstrate a superiority of our model over the state-of-the-art achieving rank-1 accuracy of 94.7% on Market-1501, 89.0% on DukeMTMC-ReID, 74.9% on CUHK03-labeled and 69.7% on CUHK03-detected.

cs.CV↗

arXiv2018

The Effects of Interfacial Recombination and Injection Barrier on the Electrical Characteristics of Perovskite Solar Cells

Charge carrier recombination in the perovskite solar cells (PSCs) has a deep influence on the electrical performance, such as open circuit voltage, short circuit current, fill factor and ultimately power conversion efficiency. The impacts of injection barrier, recombination channels, doping properties of carrier transport layers and light intensity on the performance of PSCs are theoretically investigated by drift-diffusion model in this work. The results indicate that due to the injection barrier at the interfaces of perovskite and carrier transport layer, the accumulated carriers modify the electric field distribution throughout the PSCs. Thus, a zero electric field is generated at a specific applied voltage, with greatly increases the interfacial recombination, resulting in a local kink of current density-voltage (J-V) curve. This work provides an effective strategy to improve the efficiency of PSCs by pertinently reducing both the injection barrier and interfacial recombination.

cond-mat.mes-hall↗

arXiv2017

Clogging of soft particles in 2D hoppers

Using experiments and simulations, we study the flow of soft particles through quasi-two-dimensional hoppers. The first experiment uses oil-in-water emulsion droplets in a thin sample chamber. Due to surfactants coating the droplets, they easily slide past each other, approximating soft frictionless disks. For these droplets, clogging at the hopper exit requires a narrow hopper opening only slightly larger than the droplet diameter. The second experiments use soft hydrogel particles in a thin sample chamber, where we vary gravity by changing the tilt angle of the chamber. For reduced gravity, clogging becomes easier, and can occur for larger hopper openings. Our simulations mimic the emulsion experiments and demonstrate that softness is a key factor controlling clogging: with stiffer particles or a weaker gravitational force, clogging is easier. The fractional amount a single particle is deformed under its own weight is a useful parameter measuring particle softness. Data from the simulation and hydrogel experiments collapse when compared using this parameter. Our results suggest that prior studies using hard particles were in a limit where the role of softness is negligible which causes clogging to occur with significantly larger openings.

cond-mat.soft↗

arXiv2017

Multiplicity one for the mod $p$ cohomology of Shimura curves: the tame case

Let $F$ be a totally real field, $\mathfrak{p}$ an unramified place of $F$ dividing $p$ and $\overline{r}: \mathrm{Gal}(\overline{F}/F)\rightarrow\mathrm{GL}_2(\overline{\mathbb{F}}_p)$ a continuous irreducible modular representation. The work of Buzzard, Diamond and Jarvis associates to $\overline{r}$ an admissible smooth representation of $\mathrm{GL}_2(F_\mathfrak{p})$ on the mod $p$ cohomology of Shimura curves attached to indefinite division algebras which split at $\mathfrak{p}$. When $\overline{r}|_{\mathrm{Gal}(\overline{F_\mathfrak{p}}/F_\mathfrak{p})}$ is tamely ramified and generic (and under some technical assumptions), we determine the subspace of invariants of this representation under the principal congruence subgroup of level $\mathfrak{p}$. In particular, it depends only on $\overline{r}|_{\mathrm{Gal}(\overline{F_\mathfrak{p}}/F_\mathfrak{p})}$ and verifies a multiplicity one property.

math.NT↗

arXiv2014

Sur la cohomologie des compactifications de variétés de Deligne-Lusztig

In this article, we study the étale cohomology of the compactification of Deligne-Lusztig varieties associated to a Coxeter element. We prove a result for the integral coefficients in the case of general linear group $GL_d$, and we conjecture that the similar result holds for general reductive groups.

math.AG↗

arXiv2014

L'espace symétrique de Drinfeld et correspondance de Langlands locale I

Let $K$ be a finite extension of $\mathbb{Q}_p$. We study the geometry and cohomology of the tamely ramified cover of Drinfeld's symmetric space over $K.$ We prove, in a purely local way, that the cohomology in middle degree realises the local Langlands correspondence and local Jacquet-Langlands correspondence for the depth zero supercuspidal representations.

math.NT↗

arXiv2014

L'espace de Drinfeld et correspondance de Langlands locale II

We study the geometry and the cohomology of the tamely ramified cover of Drinfeld's $p$-adic symmetric space. For this tame level, we prove, in a purely local way, most of a conjecture of Harris on the form of the $\ell$-adic cohomologies, as well as the torsion freeness of the integral cohomology. During the study, we obtain some results of independent interest on the coefficient systems over the Bruhat-Tits building associated to $GL_d(K).$

math.NT↗