Search

The AC conduction of epitaxially-grown SmB$_6$ thin films and superconducting heterostructures of SmB$_6$/YB$_6$ are investigated via time domain terahertz spectroscopy. A two-channel model of thickness-dependent bulk states and thickness-independent surface states accurately describes the measured conductance of bare SmB$_6$ thin films, demonstrating the presence of surface states in SmB$_6$. While the observed reductions in the simultaneously-measured superconducting gap, transition temperature, and superfluid density of SmB$_6$/YB$_6$ heterostructures relative to bare YB$_6$ indicate the penetration of proximity-induced superconductivity into the SmB$_6$ overlayer; the corresponding SmB$_6$-thickness independence between different heterostructures indicates that the induced superconductivity is predominantly confined to the interface surface state of the SmB$_6$. This study demonstrates the ability of terahertz spectroscopy to probe proximity-induced superconductivity at an interface buried within a heterostructure, and our results show that SmB$_6$ behaves as a predominantly insulating bulk surrounded by conducting surface states in both the normal and induced-superconducting states in both terahertz and DC responses, which is consistent with the topological Kondo insulator picture., Comment: 7 pages, 3 figures. Accepted to Phys. Rev. Lett

Language allows humans to build mental models that interpret what is happening around them resulting in more accurate long-term predictions. We present a novel trajectory prediction model that uses linguistic intermediate representations to forecast trajectories, and is trained using trajectory samples with partially-annotated captions. The model learns the meaning of each of the words without direct per-word supervision. At inference time, it generates a linguistic description of trajectories which captures maneuvers and interactions over an extended time interval. This generated description is used to refine predictions of the trajectories of multiple agents. We train and validate our model on the Argoverse dataset, and demonstrate improved accuracy results in trajectory prediction. In addition, our model is more interpretable: it presents part of its reasoning in plain language as captions, which can aid model development and can aid in building confidence in the model before deploying it., Comment: Accepted in ICRA 2022

When predicting trajectories of road agents, motion predictors usually approximate the future distribution by a limited number of samples. This constraint requires the predictors to generate samples that best support the task given task specifications. However, existing predictors are often optimized and evaluated via task-agnostic measures without accounting for the use of predictions in downstream tasks, and thus could result in sub-optimal task performance. In this paper, we propose a task-informed motion prediction model that better supports the tasks through its predictions, by jointly reasoning about prediction accuracy and the utility of the downstream tasks, which is commonly used to evaluate the task performance. The task utility function does not require the full task information, but rather a specification of the utility of the task, resulting in predictors that serve a wide range of downstream tasks. We demonstrate our approach on two use cases of common decision making tasks and their utility functions, in the context of autonomous driving and parallel autonomy. Experiment results show that our predictor produces accurate predictions that improve the task performance by a large margin in both tasks when compared to task-agnostic baselines on the Waymo Open Motion dataset., Comment: 9 pages, 5 figures, 5 tables

Modeling multi-modal high-level intent is important for ensuring diversity in trajectory prediction. Existing approaches explore the discrete nature of human intent before predicting continuous trajectories, to improve accuracy and support explainability. However, these approaches often assume the intent to remain fixed over the prediction horizon, which is problematic in practice, especially over longer horizons. To overcome this limitation, we introduce HYPER, a general and expressive hybrid prediction framework that models evolving human intent. By modeling traffic agents as a hybrid discrete-continuous system, our approach is capable of predicting discrete intent changes over time. We learn the probabilistic hybrid model via a maximum likelihood estimation problem and leverage neural proposal distributions to sample adaptively from the exponentially growing discrete space. The overall approach affords a better trade-off between accuracy and coverage. We train and validate our model on the Argoverse dataset, and demonstrate its effectiveness through comprehensive ablation studies and comparisons with state-of-the-art models., Comment: 12 pages, 10 figures, 4 tables

LaVO$_3$ (LVO) has been proposed as a promising material for photovoltaics because its strongly correlated \textit{d} electrons can facilitate the creation of multiple electron-hole pairs per incoming photon, which would lead to increased device efficiency. In this study, we intentionally grow off-stoichiometric LVO films by changing the growth conditions such as laser fluence. Our aim is to study how deviating La:V stoichiometries affect the electronic properties of LVO thin films. We find that the off-stoichiometry clearly alters the physical properties of the films. Structural characterization shows that both La-rich and V-rich films have different levels of structural distortion, with La-rich (V-rich) films showing a larger (smaller) out-of-plane lattice parameter compared to what one would expect from epitaxial strain effects alone. Both types of films show deviation from the behavior of bulk LVO in optical measurement, i.e., they do not show signatures of the expected long range orbital order, which can be a result of the structural distortions or the presence of structural domains. In transport measurements, La-rich films display clear signatures of electronic phase separation accompanying a temperature induced metal-insulator transition, while V-rich films behave as Mott insulators. The out-of-plane lattice parameter plays a crucial role in determining the transport properties, as the crossover from Mott-insulating to disorder-induced phase-separated behavior occurs around a lattice parameter value of 3.96 $\overset{\circ}{\mathrm{A}}$, quite different from what has been previously reported., Comment: Main text: 8 pages, 9 figures; Supplementary materials: 5 pages, 7 figures

Predicting driver intentions is a difficult and crucial task for advanced driver assistance systems. Traditional confidence measures on predictions often ignore the way predicted trajectories affect downstream decisions for safe driving. In this paper, we propose a novel multi-task intent recognition neural network that predicts not only probabilistic driver trajectories, but also utility statistics associated with the predictions for a given downstream task. We establish a decision criterion for parallel autonomy that takes into account the role of driver trajectory prediction in real-time decision making by reasoning about estimated task-specific utility statistics. We further improve the robustness of our system by considering uncertainties in downstream planning tasks that may lead to unsafe decisions. We test our online system on a realistic urban driving dataset, and demonstrate its advantage in terms of recall and fall-out metrics compared to baseline methods, and demonstrate its effectiveness in intervention and warning use cases., Comment: Accepted at ICRA'21/RA-L'21. Author version with 9 pages, 5 figures, 2 algorithms

Vehicle trajectory prediction is crucial for autonomous driving and advanced driver assistant systems. While existing approaches may sample from a predicted distribution of vehicle trajectories, they lack the ability to explore it -- a key ability for evaluating safety from a planning and verification perspective. In this work, we devise a novel approach for generating realistic and diverse vehicle trajectories. We extend the generative adversarial network (GAN) framework with a low-dimensional approximate semantic space, and shape that space to capture semantics such as merging and turning. We sample from this space in a way that mimics the predicted distribution, but allows us to control coverage of semantically distinct outcomes. We validate our approach on a publicly available dataset and show results that achieve state-of-the-art prediction performance, while providing improved coverage of the space of predicted trajectory semantics., Comment: 8 pages, 5 figures, 1 table

Spin-forbidden intravalley dark exciton in tungsten-based transition metal dichalcogenides (TMDCs), owing to its unique spin texture and long lifetime, has attracted intense research interest. Here, we show that we can control the dark exciton electrostatically by dressing it with one free electron or free hole, forming the dark trions. The existence of the dark trions is suggested by the unique magneto-photoluminescence spectroscopy pattern of the boron nitride (BN) encapsulated monolayer WSe2 device at low temperature. The unambiguous evidence of the dark trions is further obtained by directly resolving the radiation pattern of the dark trions through back focal plane imaging. The dark trions possess binding energy of ~ 15 meV, and it inherits the long lifetime and large g-factor from the dark exciton. Interestingly, under the out-of-plane magnetic field, dressing the dark exciton with one free electron or hole results in distinctively different valley polarization of the emitted phonon, a result of the different intervalley scattering mechanism for the electron and hole. Finally, the lifetime of the positive dark trion can be further tuned from ~ 50 to ~ 215 ps by controlling the gate voltage. The gate tunable dark trions ushers in new opportunities for excitonic optoelectronics and valleytronics.

Predicting the motion of a driver's vehicle is crucial for advanced driving systems, enabling detection of potential risks towards shared control between the driver and automation systems. In this paper, we propose a variational neural network approach that predicts future driver trajectory distributions for the vehicle based on multiple sensors. Our predictor generates both a conditional variational distribution of future trajectories, as well as a confidence estimate for different time horizons. Our approach allows us to handle inherently uncertain situations, and reason about information gain from each input, as well as combine our model with additional predictors, creating a mixture of experts. We show how to augment the variational predictor with a physics-based predictor, and based on their confidence estimations, improve overall system performance. The resulting combined model is aware of the uncertainty associated with its predictions, which can help the vehicle autonomy to make decisions with more confidence. The model is validated on real-world urban driving data collected in multiple locations. This validation demonstrates that our approach improves the prediction error of a physics-based model by 25% while successfully identifying the uncertain cases with 82% accuracy., Comment: Accepted at ICRA'19. 8 pages, 9 figures, 1 table. Video at https://youtu.be/clR08hRdtlM

We fabricated ambipolar field-effect transistors (FETs) from multi-layered triclinic ReSe2, mechanically exfoliated onto a SiO2 layer grown on p-doped Si. In contrast to previous reports on thin layers (~2 to 3 layers), we extract field-effect carrier mobilities in excess of 10^2 cm^2/Vs at room temperature in crystals with nearly ~10 atomic layers. These thicker FETs also show nearly zero threshold gate voltage for conduction and high ON to OFF current ratios when compared to the FETs built from thinner layers. We also demonstrate that it is possible to utilize this ambipolarity to fabricate logical elements or digital synthesizers. For instance, we demonstrate that one can produce simple, gate-voltage tunable phase modulators with the ability to shift the phase of the input signal by either 90^o or nearly 180^o. Given that it is possible to engineer these same elements with improved architectures, for example on h-BN in order to decrease the threshold gate voltage and increase the carrier mobilities, it is possible to improve their characteristics in order to engineer ultra-thin layered logic elements based on ReSe2., Comment: Scientific Reports 2018

We report on a study of the inelastic scattering properties of (001) and (111) Ho$_2$Ti$_2$O$_7$ single crystals at room temperature. Structural and compositional analysis along with absorption measurement confirms single crystalline phase of all samples. Room temperature polarized Raman measurements were performed on crystals in non-resonant and two different resonant conditions by using six different laser excitation lines. Lorentzian model fitting analysis is performed on all measured spectra in order to identify the difference in the Raman scattering cross-section in resonant and non-resonant conditions. Variations in the fitting parameters on account of different polarization configurations and crystallographic orientations has helped identifying their symmetry if present. Several possible scattering pathways are discussed in order to qualitatively explain the anomalous scattering results in Ho$_2$Ti$_2$O$_7$

Recently, two-dimensional materials and in particular transition metal dichalcogenides (TMDs) were extensively studied because of their strong light-matter interaction and the remarkable optoelectronic response of their field-effect transistors (FETs). Here, we report a photoconductivity study from FETs built from few-layers of p-WSe2 measured in a multi-terminal configuration under illumination by a 532 nm laser source. The photogenerated current was measured as a function of the incident optical power, of the drain-to-source bias and of the gate voltage. We observe a considerably larger photoconductivity when the phototransistors were measured via a four-terminal configuration when compared to a two-terminal one. For an incident laser power of 248 nW, we extract 18 A/W and ~4000% for the two-terminal responsivity (R) and the concomitant external quantum efficiency (EQE) respectively, when a bias voltage Vds = 1 V and a gate voltage Vbg = 10 V are applied to the sample. R and EQE are observed to increase by 370% to ~85 A/W and ~20000% respectively, when using a four-terminal configuration. Thus, we conclude that previous reports have severely underestimated the optoelectronic response of transition metal dichalcogenides, which in fact reveals a remarkable potential for photosensing applications., Comment: 6 Figures and Supporting Informations, (Accepted in "2D Materials")

The origins of indirect spin exchange in crystalline thin films of Copper Octabutoxy Phthalocyanine (Cu-OBPc) are investigated using Magnetic Circular Dichroism (MCD) spectroscopy. These studies are made possible by a solution deposition technique which produces highly ordered films with macroscopic grain sizes suitable for optical studies. For temperatures lower than 2 K, the contribution of a specific state in the valence band manifold originating from the hybridized lone pair in nitrogen orbitals of the Phthalocyanine ring, bears the Brillouin-like signature of an exchange interaction with the localized $\textit{d}$-shell Cu spins. A comprehensive MCD spectral analysis coupled with a molecular field model of a $\sigma\pi-d$ exchange analogous to $\textit{sp-d}$ interactions in Diluted Magnetic Semiconductors (DMS) renders an enhanced Zeeman splitting and a modified $\textit{g}$-factor of -4 for the electrons that mediate the interaction. These studies define an experimental tool for identifying electronic states involved in spin-dependent exchange interactions in organic materials., Comment: under review with Nature:Scientific Reports

Here, we report the photoconducting response of field-effect transistors based on three atomic layers of chemical vapor transport grown WSe$_2$ crystals mechanically exfoliated onto SiO$_2$. We find that tri-layered WSe$_2$ field-effect transistors, built with the simplest possible architecture, can display high hole mobilities ranging from 350 cm$^2$/Vs at room temperature (saturating at a value of ~500 cm$^2$/Vs below 50 K) displaying a strong photocurrent response which leads to exceptionally high photo responsivities up to 7 A/W under white light illumination of the entire channel for power densities p < 10$^2$ W/m$^2$. Under a fixed wavelength of $\lambda$ = 532 nm and a laser spot size smaller than the conducting channel area we extract photo responsitivities approaching 100 mA/W with concomitantly high external quantum efficiencies up to ~ 40 % at room temperature. These values surpass values recently reported from more complex architectures, such as graphene and transition metal dichalcogenides based heterostructures. Also, tri-layered WSe$_2$ photo-transistors display photo response times in the order of 10 microseconds. Our results indicate that the addition of a few atomic layers considerably decreases the photo response times, probably by minimizing the interaction with the substrates, while maintaining a very high photo-responsivity., Comment: 39 pages, 6 figures, and supplemental information including 5 supplemental figures. To appear in ACS Applied Materials and Interfaces