arXiv subjects

Yu Liu

Publications and source records attributed to Yu Liu.

At least 775 records · Page 43Linked to original sources

arXiv2022

DiffGAR: Model-Agnostic Restoration from Generative Artifacts Using Image-to-Image Diffusion Models

Recent generative models show impressive results in photo-realistic image generation. However, artifacts often inevitably appear in the generated results, leading to downgraded user experience and reduced performance in downstream tasks. This work aims to develop a plugin post-processing module for diverse generative models, which can faithfully restore images from diverse generative artifacts. This is challenging because: (1) Unlike traditional degradation patterns, generative artifacts are non-linear and the transformation function is highly complex. (2) There are no readily available artifact-image pairs. (3) Different from model-specific anti-artifact methods, a model-agnostic framework views the generator as a black-box machine and has no access to the architecture details. In this work, we first design a group of mechanisms to simulate generative artifacts of popular generators (i.e., GANs, autoregressive models, and diffusion models), given real images. Second, we implement the model-agnostic anti-artifact framework as an image-to-image diffusion model, due to its advantage in generation quality and capacity. Finally, we design a conditioning scheme for the diffusion model to enable both blind and non-blind image restoration. A guidance parameter is also introduced to allow for a trade-off between restoration accuracy and image quality. Extensive experiments show that our method significantly outperforms previous approaches on the proposed datasets and real-world artifact images.

cs.CV↗

arXiv2022

Enhanced Superconductivity and Electron Correlations in Intercalated ZrTe$_3$

Charge density waves (CDWs) with superconductivity, competing Fermi surface instabilities and collective orders, have captured much interest in two-dimensional van der Waals (vdW) materials. Understanding of CDW suppression mechanism, its connection to emerging superconducting state and electronic correlations provides opportunities for engineering the electronic properties of vdW heterostructures and thin film devices. Using combination of the thermal transport, X-ray photoemission spectroscopy, Raman measurements, and first-principle calculations, we observe an increase in electronic correlations of the conducting states as CDW is suppressed in ZrTe$_3$ with 5\% Cu and Ni intercalation in the vdW gap. As superconductivity emerges, intercalation brings decoupling of quasi-one-dimensional conduction electrons with phonons as a consequence of intercalation-induced lattice expansion but also a drastic increase in Zr$^{2+}$ at the expense of Zr$^{4+}$ metal atoms. These observation demonstrate the potential of atomic intercalates in vdW gap for ground state tuning but also illustrate the crucial role of Zr metal valence in formation of collective electronic orders.

cond-mat.supr-con↗

arXiv2022

Physical properties of the layered $f$-electron van der Waals magnet Ce$_2$Te$_5$

We report a detailed study of the magnetic, transport, and thermodynamic properties of Ce$_2$Te$_5$ single crystals, a layered $f$-electron van der Waals magnet. Four consecutive transitions at $\sim$ 5.2, 2.1, 0.9, and 0.4 K were observed in the $ac$-plane electrical resistivity $ρ$(T), which were further confirmed in specific heat $C_\textrm{p}$(T) measurements. Analysis of the magnetic susceptibility $χ$(T), the magnetic-field variation of $ρ$(T), and the increase of the first transition temperature ($T_\textrm{c} \sim$ 5.2 K) with applied magnetic field indicates ferromagnetic order, while the decrease of the other transitions with field suggests different states with dominant antiferromagnetic interactions below $T_2 \sim$ 2.1 K, $T_3 \sim$ 0.9 K, and $T_4$ = 0.4 K. Critical behavior analysis around $T_\textrm{c}$ that gives critical exponents $β= 0.31(2)$, $γ= 0.99(2)$, $δ= 4.46(1)$, $T_\textrm{c} = 5.32(1)$ K indicates that Ce$_2$Te$_5$ shows a three-dimensional magnetic critical behavior. Moreover, the Hall resistivity $ρ_{\textrm{xy}}$ indicates that Ce$_2$Te$_5$ is a multi-band system with a relatively high electron mobility $\sim 2900$ cm$^2$ V$^{-1}$ s$^{-1}$ near $T_\textrm{c}$, providing further opportunities for future device applications.

cond-mat.str-el↗

arXiv2022

Gate-tunable kinetic inductance in proximitized nanowires

We report the detection of a gate-tunable kinetic inductance in a hybrid InAs/Al nanowire. For this purpose, we have embedded the nanowire into a quarter-wave coplanar waveguide resonator and measured the resonance frequency of the circuit. We find that the resonance frequency can be changed via the gate voltage that controls the electron density of the proximitized semiconductor and thus the nanowire inductance. Applying Mattis-Bardeen theory, we extract the gate dependence of the normal state conductivity of the nanowire, as well as its superconducting gap. Our measurements complement existing characterization methods for hybrid nanowires and provide a new and useful tool for gate-controlled superconducting electronics.

cond-mat.supr-con↗

arXiv2022

Interplay of broken symmetry and delocalized excitations in the insulating state of 1$T$-TaS$_2$

Coexistence of localized and extended excitations is central to the macroscopic properties of correlated materials. For 5d transition metal compounds, electron correlations alone generally do not lead to a metal-insulator (Mott) transition, with insulating behavior usually resulting from their coupling with magnetic ordering and/or structural distortions. 1$T$-TaS$_2$ is a prototypical example of such correlated insulating behavior, with a high-symmetry metallic phase transforming into a distorted, charge density wave (CDW) insulating state at low temperatures. The relevance of the localized electron physics at play in 3d compounds to these 5d transition metal compounds remains an open question. We resolved this standing controversy in 1$T$-TaS$_2$ combining resonant inelastic X-ray spectroscopy and first-principles calculations. We observed five electronic excitations arising from the interband transitions of the Ta 5d orbitals and the S 3p ligand state, with none of the excitations on the order of the Mott gap. These excitations cannot be explained within the framework of standard multiplet calculations that assume a localized wavefunction, but instead, are captured by a band theory framework accounting for the low symmetry of the crystal field in the CDW state. Our findings suggest that the electronic property of 1$T$-TaS$_2$ is dominated by both plasmonic quasiparticles and inter-band transitions associated with a Drude-type response, with no resonance associated with a putative Mott transition. Our discovery provides new insights into the electron localization and the onset of insulating behavior in 5d transition metal materials.

cond-mat.mtrl-sci↗

arXiv2022

Quantum capacitance of a superconducting subgap state in an electrostatically floating dot-island

We study a hybrid device defined in an InAs nanowire with an epitaxial Al shell that consists of a quantum dot in contact with a superconducting island. The device is electrically floating, prohibiting transport measurements, but providing access to states that would otherwise be highly excited and unstable. Radio-frequency reflectometry with lumped-element resonators couples capacitatively to the quantum dot, and detects the presence of discrete subgap states. We perform a detailed study of the case with no island states, but with quantum-dot-induced subgap states controlled by the tunnel coupling. When the gap to the quasi-continuum of the excited states is small, the capacitance loading the resonator is strongly suppressed by thermal excitations, an effect we dub "thermal screening". The resonance frequency shift and changes in the quality factor at charge transitions can be accounted for using a single-level Anderson impurity model. The established measurement method, as well as the analysis and simulation framework, are applicable to more complex hybrid devices such as Andreev molecules or Kitaev chains.

cond-mat.mes-hall↗

arXiv2022

Long range order for three-dimensional random field Ising model throughout the entire low temperature regime

For $d\geq 3$, we study the Ising model on $\mathbb Z^d$ with random field given by $\{εh_v: v\in \mathbb Z^d\}$ where $h_v$'s are independent normal variables with mean 0 and variance 1. We show that for any $T < T_c$ (here $T_c$ is the critical temperature without disorder), long range order exists as long as $ε$ is sufficiently small depending on $T$. Our work extends previous results of Imbrie (1985) and Bricmont--Kupiainen (1988) from the very low temperature regime to the entire low temperature regime.

math.PR↗

arXiv2022

Accelerating Reinforcement Learning for Autonomous Driving using Task-Agnostic and Ego-Centric Motion Skills

Efficient and effective exploration in continuous space is a central problem in applying reinforcement learning (RL) to autonomous driving. Skills learned from expert demonstrations or designed for specific tasks can benefit the exploration, but they are usually costly-collected, unbalanced/sub-optimal, or failing to transfer to diverse tasks. However, human drivers can adapt to varied driving tasks without demonstrations by taking efficient and structural explorations in the entire skill space rather than a limited space with task-specific skills. Inspired by the above fact, we propose an RL algorithm exploring all feasible motion skills instead of a limited set of task-specific and object-centric skills. Without demonstrations, our method can still perform well in diverse tasks. First, we build a task-agnostic and ego-centric (TaEc) motion skill library in a pure motion perspective, which is diverse enough to be reusable in different complex tasks. The motion skills are then encoded into a low-dimension latent skill space, in which RL can do exploration efficiently. Validations in various challenging driving scenarios demonstrate that our proposed method, TaEc-RL, outperforms its counterparts significantly in learning efficiency and task performance.

cs.RO↗

arXiv2022

Identification of the anomalous fast bulk events in a p-type point contact germanium detector

The ultralow detection threshold, ultralow intrinsic background, and excellent energy resolution of p-type point-contact germanium detector are important for rare-event searches, in particular for the detection of direct dark matter interactions, coherent elastic neutrino-nucleus scattering, and neutrinoless double beta decay. Anomalous bulk events with an extremely fast rise time are observed in the CDEX-1B detector. We report a method of extracting fast bulk events from bulk events using a pulse shape simulation and reconstructed source experiment signature. Calibration data and the distribution of X-rays generated by intrinsic radioactivity verified that the fast bulk experienced a single hit near the passivation layer. The performance of this germanium detector indicates that it is capable of single-hit bulk spatial resolution and thus provides a background removal technique.

physics.ins-det↗

arXiv2022

Inline Tests

Unit tests are widely used to check source code quality, but they can be too coarse-grained or ill-suited for testing individual program statements. We introduce inline tests to make it easier to check for faults in statements. We motivate inline tests through several language features and a common testing scenario in which inline tests could be beneficial. For example, inline tests can allow a developer to test a regular expression in place. We also define language-agnostic requirements for inline testing frameworks. Lastly, we implement I-Test, the first inline testing framework. I-Test works for Python and Java, and it satisfies most of the requirements. We evaluate I-Test on open-source projects by using it to test 144 statements in 31 Python programs and 37 Java programs. We also perform a user study. All nine user study participants say that inline tests are easy to write and that inline testing is beneficial. The cost of running inline tests is negligible, at 0.007x--0.014x, and our inline tests helped find two faults that have been fixed by the developers.

cs.SE↗

arXiv2022

Self-slimmed Vision Transformer

Vision transformers (ViTs) have become the popular structures and outperformed convolutional neural networks (CNNs) on various vision tasks. However, such powerful transformers bring a huge computation burden, because of the exhausting token-to-token comparison. The previous works focus on dropping insignificant tokens to reduce the computational cost of ViTs. But when the dropping ratio increases, this hard manner will inevitably discard the vital tokens, which limits its efficiency. To solve the issue, we propose a generic self-slimmed learning approach for vanilla ViTs, namely SiT. Specifically, we first design a novel Token Slimming Module (TSM), which can boost the inference efficiency of ViTs by dynamic token aggregation. As a general method of token hard dropping, our TSM softly integrates redundant tokens into fewer informative ones. It can dynamically zoom visual attention without cutting off discriminative token relations in the images, even with a high slimming ratio. Furthermore, we introduce a concise Feature Recalibration Distillation (FRD) framework, wherein we design a reverse version of TSM (RTSM) to recalibrate the unstructured token in a flexible auto-encoder manner. Due to the similar structure between teacher and student, our FRD can effectively leverage structure knowledge for better convergence. Finally, we conduct extensive experiments to evaluate our SiT. It demonstrates that our method can speed up ViTs by 1.7x with negligible accuracy drop, and even speed up ViTs by 3.6x while maintaining 97% of their performance. Surprisingly, by simply arming LV-ViT with our SiT, we achieve new state-of-the-art performance on ImageNet. Code is available at https://github.com/Sense-X/SiT.

cs.CV↗

arXiv2022

UniNet: Unified Architecture Search with Convolution, Transformer, and MLP

Recently, transformer and multi-layer perceptron (MLP) architectures have achieved impressive results on various vision tasks. However, how to effectively combine those operators to form high-performance hybrid visual architectures still remains a challenge. In this work, we study the learnable combination of convolution, transformer, and MLP by proposing a novel unified architecture search approach. Our approach contains two key designs to achieve the search for high-performance networks. First, we model the very different searchable operators in a unified form, and thus enable the operators to be characterized with the same set of configuration parameters. In this way, the overall search space size is significantly reduced, and the total search cost becomes affordable. Second, we propose context-aware downsampling modules (DSMs) to mitigate the gap between the different types of operators. Our proposed DSMs are able to better adapt features from different types of operators, which is important for identifying high-performance hybrid architectures. Finally, we integrate configurable operators and DSMs into a unified search space and search with a Reinforcement Learning-based search algorithm to fully explore the optimal combination of the operators. To this end, we search a baseline network and scale it up to obtain a family of models, named UniNets, which achieve much better accuracy and efficiency than previous ConvNets and Transformers. In particular, our UniNet-B5 achieves 84.9% top-1 accuracy on ImageNet, outperforming EfficientNet-B7 and BoTNet-T7 with 44% and 55% fewer FLOPs respectively. By pretraining on the ImageNet-21K, our UniNet-B6 achieves 87.4%, outperforming Swin-L with 51% fewer FLOPs and 41% fewer parameters. Code is available at https://github.com/Sense-X/UniNet.

cs.CV↗

arXiv2022

Unifying Visual Perception by Dispersible Points Learning

We present a conceptually simple, flexible, and universal visual perception head for variant visual tasks, e.g., classification, object detection, instance segmentation and pose estimation, and different frameworks, such as one-stage or two-stage pipelines. Our approach effectively identifies an object in an image while simultaneously generating a high-quality bounding box or contour-based segmentation mask or set of keypoints. The method, called UniHead, views different visual perception tasks as the dispersible points learning via the transformer encoder architecture. Given a fixed spatial coordinate, UniHead adaptively scatters it to different spatial points and reasons about their relations by transformer encoder. It directly outputs the final set of predictions in the form of multiple points, allowing us to perform different visual tasks in different frameworks with the same head design. We show extensive evaluations on ImageNet classification and all three tracks of the COCO suite of challenges, including object detection, instance segmentation and pose estimation. Without bells and whistles, UniHead can unify these visual tasks via a single visual head design and achieve comparable performance compared to expert models developed for each task.We hope our simple and universal UniHead will serve as a solid baseline and help promote universal visual perception research. Code and models are available at https://github.com/Sense-X/UniHead.

cs.CV↗

arXiv2022

Towards Robust Face Recognition with Comprehensive Search

Data cleaning, architecture, and loss function design are important factors contributing to high-performance face recognition. Previously, the research community tries to improve the performance of each single aspect but failed to present a unified solution on the joint search of the optimal designs for all three aspects. In this paper, we for the first time identify that these aspects are tightly coupled to each other. Optimizing the design of each aspect actually greatly limits the performance and biases the algorithmic design. Specifically, we find that the optimal model architecture or loss function is closely coupled with the data cleaning. To eliminate the bias of single-aspect research and provide an overall understanding of the face recognition model design, we first carefully design the search space for each aspect, then a comprehensive search method is introduced to jointly search optimal data cleaning, architecture, and loss function design. In our framework, we make the proposed comprehensive search as flexible as possible, by using an innovative reinforcement learning based approach. Extensive experiments on million-level face recognition benchmarks demonstrate the effectiveness of our newly-designed search space for each aspect and the comprehensive search. We outperform expert algorithms developed for each single research track by large margins. More importantly, we analyze the difference between our searched optimal design and the independent design of the single factors. We point out that strong models tend to optimize with more difficult training datasets and loss functions. Our empirical study can provide guidance in future research towards more robust face recognition systems.

cs.CV↗

arXiv2022

Spectroscopy of spin-split Andreev levels in a quantum dot with superconducting leads

We use a hybrid superconductor-semiconductor transmon device to perform spectroscopy of a quantum dot Josephson junction tuned to be in a spin-1/2 ground state with an unpaired quasiparticle. Due to spin-orbit coupling, we resolve two flux-sensitive branches in the transmon spectrum, depending on the spin of the quasi-particle. A finite magnetic field shifts the two branches in energy, favoring one spin state and resulting in the anomalous Josephson effect. We demonstrate the excitation of the direct spin-flip transition using all-electrical control. Manipulation and control of the spin-flip transition enable the future implementation of charging energy protected Andreev spin qubits.

cond-mat.mes-hall↗

arXiv2022

Spin-phonon interaction and short range order in $\mathrm{Mn_3Si_{2}Te_6}$

The vibrational properties of ferrimagnetic $\mathrm{Mn_3Si_{2}Te_6}$ single crystals are investigated using Raman spectroscopy and density functional theory calculations. Eighteen Raman-active modes are identified, fourteen of which are assigned according to with the trigonal symmetry. Four additional peaks, obeying the $A_{1g}$ selection rules, are attributed to the overtones. The unconventional temperature evolution of the $A_{1g}^5$ mode self-energy suggests a competition between different short-range magnetic correlations that significantly impact the spin-phonon interaction in $\mathrm{Mn_3Si_{2}Te_6}$. The research provides a comprehensive insight to the lattice properties, studies their temperature dependence and shows the arguments for existence of competing short-range magnetic phases in $\mathrm{Mn_3Si_{2}Te_6}$.

cond-mat.mtrl-sci↗

arXiv2022

Direct manipulation of a superconducting spin qubit strongly coupled to a transmon qubit

Spin qubits in semiconductors are currently one of the most promising architectures for quantum computing. However, they face challenges in realizing multi-qubit interactions over extended distances. Superconducting spin qubits provide a promising alternative by encoding a qubit in the spin degree of freedom of an Andreev level. Such an Andreev spin qubit could leverage the advantages of circuit quantum electrodynamic, enabled by an intrinsic spin-supercurrent coupling. The first realization of an Andreev spin qubit encoded the qubit in the excited states of a semiconducting weak-link, leading to frequent decay out of the computational subspace. Additionally, rapid qubit manipulation was hindered by the need for indirect Raman transitions. Here, we exploit a different qubit subspace, using the spin-split doublet ground state of an electrostatically-defined quantum dot Josephson junction with large charging energy. Additionally, we use a magnetic field to enable direct spin manipulation over a frequency range of 10 GHz. Using an all-electric microwave drive we achieve Rabi frequencies exceeding 200 MHz. We furthermore embed the Andreev spin qubit in a superconducting transmon qubit, demonstrating strong coherent qubit-qubit coupling. These results are a crucial step towards a hybrid architecture that combines the beneficial aspects of both superconducting and semiconductor qubits.

quant-ph↗

arXiv2022

Suppression of Superconductivity and Nematic Order in Fe$_{1-y}$Se$_{1-x}$S$_x$ (0$\leq$$x$$\leq$1, $y$$\leq$0.1) Crystals by Anion Height Disorder

Connections between crystal chemistry and critical temperature $T_c$ have been in the focus of superconductivity, one of the most widely studied phenomena in physics, chemistry and materials science alike. In most Fe-based superconductors, materials chemistry and physics conspire so that $T_c$ correlates with the average anion height above the Fe plane, i. e. with the geometry of the FeAs4 or FeCh4 (Ch = Te, Se, or S) tetrahedron. By synthesizing Fe$_{1-y}$Se$_{1-x}$S$_x$ (0$\leq$$x$$\leq$1, $y$$\leq$0.1), we find that in alloyed crystals $T_c$ is not correlated with the anion height like it is for most other Fe superconductors. Instead, changes in $T_c$($x$) and tetragonal-to-orthorombic (nematic) transition $T_s$($x$) upon cooling are correlated with disorder in Fe vibrations in the direction orthogonal to Fe planes, along the crystallographic c-axis. The disorder stems from the random nature of S substitution, causing deformed Fe(Se,S)4 tetrahedra with different Fe-Se and Fe-S bond distances. Our results provide evidence of $T_c$ and $T_s$ suppression by disorder in anion height. The connection to local crystal chemistry may be exploited in computational prediction of new superconducting materials with Fe/S building blocks.

cond-mat.supr-con↗