arXiv subjects

Weijie Li

Publications and source records attributed to Weijie Li.

51 records · Page 3Linked to original sources

arXiv2023

Charge states, triple points and quadruple points in an InAs nanowire triple quantum dot revealed by an integrated charge sensor

A serial triple quantum dot (TQD) integrated with a quantum dot (QD) charge sensor is realized from an InAs nanowire via a fine finger-gate technique. The complex charge states and intriguing properties of the device are studied in the few-electron regime by direct transport measurements and by charge-sensor detection measurements. The measurements of the charge stability diagram for a capacitively coupled, parallel double-QD formed from a QD in the TQD and the sensor QD show a visible capacitance coupling between the TQD and the sensor QD, indicating a good sensitivity of the charge sensor. The charge stability diagrams of the TQD are measured by the charge sensor and the global features seen in the measured charge stability diagrams are well reproduced by the simultaneous measurements of the direct transport current through the TQD and by the simulation made based on an effective capacitance network model. The complex charge stability diagrams of the TQD are measured in detail with the integrated charge sensor in an energetically degenerate region, where all the three QDs are on or nearly on resonance, and the formations of quadruple points and of all possible eight charge states are observed. In addition, the operation of the TQD as a quantum cellular automata is demonstrated and discussed.

cond-mat.mes-hall↗

arXiv2022

STAR: SQL Guided Pre-Training for Context-dependent Text-to-SQL Parsing

In this paper, we propose a novel SQL guided pre-training framework STAR for context-dependent text-to-SQL parsing, which leverages contextual information to enrich natural language (NL) utterance and table schema representations for text-to-SQL conversations. Concretely, we propose two novel pre-training objectives which respectively explore the context-dependent interactions of NL utterances and SQL queries within each text-to-SQL conversation: (i) schema state tracking (SST) objective that tracks and explores the schema states of context-dependent SQL queries in the form of schema-states by predicting and updating the value of each schema slot during interaction; (ii) utterance dependency tracking (UDT) objective that employs weighted contrastive learning to pull together two semantically similar NL utterances and push away the representations of semantically dissimilar NL utterances within each conversation. In addition, we construct a high-quality large-scale context-dependent text-to-SQL conversation corpus to pre-train STAR. Extensive experiments show that STAR achieves new state-of-the-art performance on two downstream benchmarks (SParC and CoSQL), significantly outperforming previous pre-training methods and ranking first on the leaderboard. We believe the release of the constructed corpus, codebase and pre-trained STAR checkpoints would push forward the research in this area. For reproducibility, we release our code and data at https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/star.

cs.CL↗

arXiv2022

Quadrupolar excitons in a tunnel-coupled van der Waals heterotrilayer

Strongly bound excitons and many-body interactions between them determine light-matter interactions in van der Waals (vdW) heterostructures of 2D semiconductors. Unlike fundamental particles, quasiparticles in condensed matter, such as excitons, can be tailored to alter their interactions and realize emergent quantum phases. Here, using a WS$_2$/WSe$_2$/WS$_2$ heterotrilayer, we create a quantum superposition of oppositely oriented dipolar excitons - a quadrupolar exciton - wherein an electron is layer-hybridized in WS$_2$ layers while the hole localizes in WSe$_2$. In contrast to dipolar excitons, symmetric quadrupolar excitons only redshift in an out-of-plane electric field, consistent with ab initio calculations, regaining dipolar characteristics at higher fields. Electric field tunes the hybridization and allows for lifetime control through modification of the excitonic wavefunction. Lack of density-dependent blue shift of heterotrilayer excitons compared to dipolar excitons is consistent with quadrupolar interactions. Our results present vdW heterotrilayers as a field-tunable platform to engineer light-matter interactions and explore quantum phase transitions between spontaneously ordered many-exciton phases.

cond-mat.mes-hall↗

arXiv2022

A Dynamic Subarray Structure in Reconfigurable Intelligent Surfaces for TeraHertz Communication Systems

Reconfigurable Intelligent Surface (RIS) has become a popular technology to improve the capability of a THz multiuser Multi-input multi-output (MIMO) communication system. THz wave characteristics, on the other hand, restrict THz beam coverage on RIS when using a uniform planar array (UPA) antenna. In this study, we propose a dynamic RIS subarray structure to improve the performance of a THz MIMO communication system. In more details, an RIS is divided into several RIS subarrays according to the number of users. Each RIS subarray is paired with a user and only reflects beams to the corresponding user. Based on the structure of RIS, we first propose a weighted minimum mean square error - RIS local search (WMMSE-LS) scheme, which requires that each RIS element has limited phase shifts. To improve the joint beamforming performance, we further develop an adaptive Block Coordinate Descent(BCD)-aided algorithm, an iterative optimization method. Numerical results demonstrate the effectiveness of the dynamic RIS subarray structure and the adaptive BCD-aided joint beamforming scheme and also show the merit of our proposed system.

eess.SP↗

arXiv2022

Kinematic Motion Retargeting via Neural Latent Optimization for Learning Sign Language

Motion retargeting from a human demonstration to a robot is an effective way to reduce the professional requirements and workload of robot programming, but faces the challenges resulting from the differences between humans and robots. Traditional optimization-based methods are time-consuming and rely heavily on good initialization, while recent studies using feedforward neural networks suffer from poor generalization to unseen motions. Moreover, they neglect the topological information in human skeletons and robot structures. In this paper, we propose a novel neural latent optimization approach to address these problems. Latent optimization utilizes a decoder to establish a mapping between the latent space and the robot motion space. Afterward, the retargeting results that satisfy robot constraints can be obtained by searching for the optimal latent vector. Alongside with latent optimization, neural initialization exploits an encoder to provide a better initialization for faster and better convergence of optimization. Both the human skeleton and the robot structure are modeled as graphs to make better use of topological information. We perform experiments on retargeting Chinese sign language, which involves two arms and two hands, with additional requirements on the relative relationships among joints. Experiments include retargeting various human demonstrations to YuMi, NAO, and Pepper in the simulation environment and to YuMi in the real-world environment. Both efficiency and accuracy of the proposed method are verified.

cs.RO↗

arXiv2021

Local Sensing of Correlated Electrons in Dual-moiré Heterostructures using Dipolar Excitons

Moiré heterostructures are rapidly emerging as a tunable platform to study correlated electronic phenomena. Discovery of exotic quantum phases in moiré systems requires novel probes of charge and spin order. Unlike detection schemes which average over several moiré cells, local sensors can provide richer information with greater sensitivity. We study a WSe$_2$/MoSe$_2$/WSe$_2$ heterotrilayer which hosts excitons and electrons in distinct moiré lattices, and show that localized dipolar excitons are sensitive proximity charge sensors, uncovering numerous correlated electronic states at fractional fillings of the multi-orbital moiré lattice. In addition, the emission polarization can reveal the local electronic spin configuration at different fillings. Our results establish dipolar excitons as promising candidates to study emergent quantum matter and quantum magnetism in moiré crystals with higher spatial resolution.

cond-mat.mes-hall↗

arXiv2021

Hierarchical Object-to-Zone Graph for Object Navigation

The goal of object navigation is to reach the expected objects according to visual information in the unseen environments. Previous works usually implement deep models to train an agent to predict actions in real-time. However, in the unseen environment, when the target object is not in egocentric view, the agent may not be able to make wise decisions due to the lack of guidance. In this paper, we propose a hierarchical object-to-zone (HOZ) graph to guide the agent in a coarse-to-fine manner, and an online-learning mechanism is also proposed to update HOZ according to the real-time observation in new environments. In particular, the HOZ graph is composed of scene nodes, zone nodes and object nodes. With the pre-learned HOZ graph, the real-time observation and the target goal, the agent can constantly plan an optimal path from zone to zone. In the estimated path, the next potential zone is regarded as sub-goal, which is also fed into the deep reinforcement learning model for action prediction. Our methods are evaluated on the AI2-Thor simulator. In addition to widely used evaluation metrics SR and SPL, we also propose a new evaluation metric of SAE that focuses on the effective action rate. Experimental results demonstrate the effectiveness and efficiency of our proposed method.

cs.CV↗

arXiv2021

CORAL: Colored structural representation for bi-modal place recognition

Place recognition is indispensable for a drift-free localization system. Due to the variations of the environment, place recognition using single-modality has limitations. In this paper, we propose a bi-modal place recognition method, which can extract a compound global descriptor from the two modalities, vision and LiDAR. Specifically, we first build the elevation image generated from 3D points as a structural representation. Then, we derive the correspondences between 3D points and image pixels that are further used in merging the pixel-wise visual features into the elevation map grids. In this way, we fuse the structural features and visual features in the consistent bird-eye view frame, yielding a semantic representation, namely CORAL. And the whole network is called CORAL-VLAD. Comparisons on the Oxford RobotCar show that CORAL-VLAD has superior performance against other state-of-the-art methods. We also demonstrate that our network can be generalized to other scenes and sensor configurations on cross-city datasets.

cs.CV↗

arXiv2021

Charge detection of a quantum dot under different tunneling barrier symmetries and bias voltages

We report on the realization of a coupled quantum dot (QD) system containing two single QDs made in two adjacent InAs nanowires. One QD (sensor QD) is used as a charge sensor to detect the charge state transition in the other QD (target QD). We investigate the effect of the tunneling barrier asymmetry of the target QD on the detection visibility of charge state transition in the target QD. The charge stability diagrams of the target QD under different configurations of barrier-gate voltages are simultaneously measured via the direct signals of electron transport through the target QD and via the detection signals of charge state transition in the target QD from the sensor QD. We find that the complete Coulomb diamond boundaries of the target QD and the transport processes involving the excited states in the target QD can be observed in the transconductance signals of the sensor QD only when the tunneling barriers of the target QD are nearly symmetric. These phenomena are explained by analyzing the effect of the ratio of the two tunneling rates on the electron transport processes through the target QD. Our results imply that it is important to consider the symmetry of the tunneling couplings when constructing a charge sensor integrated QD device or qubit.

cond-mat.mes-hall↗

arXiv2021

Dynamic Movement Primitive based Motion Retargeting for Dual-Arm Sign Language Motions

We aim to develop an efficient programming method for equipping service robots with the skill of performing sign language motions. This paper addresses the problem of transferring complex dual-arm sign language motions characterized by the coordination among arms and hands from human to robot, which is seldom considered in previous studies of motion retargeting techniques. In this paper, we propose a novel motion retargeting method that leverages graph optimization and Dynamic Movement Primitives (DMPs) for this problem. We employ DMPs in a leader-follower manner to parameterize the original trajectories while preserving motion rhythm and relative movements between human body parts, and adopt a three-step optimization procedure to find deformed trajectories for robot motion planning while ensuring feasibility for robot execution. Experimental results of several Chinese Sign Language (CSL) motions have been successfully performed on ABB's YuMi dual-arm collaborative robot (14-DOF) with two 6-DOF Inspire-Robotics' multi-fingered hands, a system with 26 DOFs in total.

cs.RO↗

arXiv2021

Highly tunable quadruple quantum dot in a narrow bandgap semiconductor InAs nanowire

Quantum dots (QDs) made from semiconductors are among the most promising platforms for the developments of quantum computing and simulation chips, and have advantages over other platforms in high density integration and in compatibility to the standard semiconductor chip fabrication technology. However, development of a highly tunable semiconductor multiple QD system still remains as a major challenge. Here, we demonstrate realization of a highly tunable linear quadruple QD (QQD) in a narrow bandgap semiconductor InAs nanowire with fine finger gate technique. The QQD is studied by electron transport measurements in the linear response regime. Characteristic two-dimensional charge stability diagrams containing four groups of resonant current lines of different slopes are found for the QQD. It is shown that these current lines can be individually assigned as arising from resonant electron transport through the energy levels of different QDs. Benefited from the excellent gate tunability, we also demonstrate tuning of the QQD to regimes where the energy levels of two QDs, three QDs and all the four QDs are energetically on resonance, respectively, with the fermi level of source and drain contacts. A capacitance network model is developed for the linear QQD and the simulated charge stability diagrams based on the model show good agreements with the experiments. Our work presents a solid experimental evidence that narrow bandgap semiconductor nanowires multiple QDs could be used as a versatile platform to achieve integrated qubits for quantum computing and to perform quantum simulations for complex many-body systems.

cond-mat.mes-hall↗

arXiv2020

Detection of charge states of an InAs nanowire triple quantum dot with an integrated nanowire charge sensor

A linear triple quantum dot (TQD) integrated with a quantum dot (QD) charge sensor is realized. The TQD and the charge sensor are built from two adjacent InAs nanowires by fine finger gate technique. The charge state configurations of the nanowire TQD are studied by measurements of the direct transport signals of the TQD and by detection of the charge state transitions in the TQD via the nanowire QD sensor. Excellent agreements in the charge stability diagrams of the TQD obtained by the direct transport measurements and by the charge-state transition detection measurements are achieved. It is shown that the charge stability diagrams are featured by three groups of charge state transition lines of different slopes, corresponding to the changes in the electron occupation numbers of the three individual QDs in the TQD. It is also shown that the integrated nanowire QD sensor is highly sensitive and can detect the charge state transitions in the cases where the direct transport signals of the TQD are too weak to be measurable. Tuning to a regime, where all the three QDs in the TQD are close to be on resonance with the Fermi level of the source and drain reservoirs and co-existence of triple and quadruple points becomes possible, has also been demonstrated with the help of the charge sensor in the region where the direct transport signals of the TQD are hardly visible.

cond-mat.mes-hall↗

arXiv2020

Optical control of valley Zeeman effect through many-exciton interactions

Charge carriers in two-dimensional transition metal dichalcogenides (TMDs), such as WSe$_2$, have their spin and valley-pseudospin locked into an optically-addressable index that is proposed as a basis for future information processing. The manipulation of this spin-valley index requires tuning its energy, typically through external magnetic field (B), which is cumbersome. Thus, other efficient routes like all-optical control of spin-valley index are desirable. Here, we show that many-body interactions amongst interlayer excitons in WSe$_2$/MoSe$_2$ heterobilayer induce a steady-state valley Zeeman splitting corresponding to B $\sim$ 6 Tesla. This anomalous splitting, present at incident powers as low as $μ$Ws, increases with power and enhances, suppresses or even flips the sign of a B-induced splitting. Moreover, the $\it{g}$-factor of valley Zeeman splitting can be tuned by $\sim$ 30 $\%$ with incident power. In addition to valleytronics, our results are relevant for achieving optical non-reciprocity using two-dimensional materials.

cond-mat.mes-hall↗

arXiv2019

Dipolar interactions between field-tuneable, localized emitters in van der Waals heterostructures

While photons in free space barely interact, matter can mediate interactions between them resulting in optical nonlinearities. Such interactions at the single-quantum level result in an on-site photon repulsion, crucial for photon-based quantum information processing and for realizing strongly interacting many-body states of light. Here, we report repulsive dipole-dipole interactions between electric field tuneable, localized interlayer excitons in MoSe$_2$/WSe$_2$ heterobilayer. The presence of a single, localized exciton with an out-of-plane, non-oscillating dipole moment increases the energy of the second excitation by $\sim$ 2 meV -- an order of magnitude larger than the emission linewidth and corresponding to an inter-dipole distance of $\sim$ 5 nm. At higher excitation power, multi-exciton complexes appear at systematically higher energies. The magnetic field dependence of the emission polarization is consistent with spin-valley singlet nature of the dipolar molecular state. Our finding is an important step towards the creation of excitonic few- and many-body states such as dipolar crystals with spin-valley spinor in van der Waals (vdW) heterostructures.

cond-mat.mes-hall↗

arXiv2018

Magnetic freezing transition in a CoO/Permalloy bilayer revealed by transverse ac susceptibility

We utilize variable-temperature, variable-frequency magneto-optical transverse magnetic susceptibility technique to study the static and dynamical magnetic properties of a thin-film CoO/Permalloy bilayer. Our measurements demonstrate that in the studied system, the directional asymmetry of the hysteresis loop is associated mainly with the difference in the reversal mechanisms between the two reversed states of magnetization stabilized by the exchange-induced uniaxial anisotropy. The latter is found to be much larger than the exchange-induced unidirectional anisotropy of the ferromagnet. We also observe an abrupt variation of the frequency-dependent imaginary part of ac susceptibility near the exchange bias blocking temperature, consistent with the magnetic freezing transition inferred from the previous time-domain studies of magnetic aging in similar systems. The developed measurement approach enables precise characterization of the dynamical and static characteristics of thin-film magnetic heterostructures that can find applications in reconfigurable magnonic and neuromorphic circuits.

cond-mat.mtrl-sci↗