arXiv subjects

Haoxiang Li

Publications and source records attributed to Haoxiang Li.

At least 73 records · Page 4Linked to original sources

arXiv2021

Experimental Observation of Efficient Nonreciprocal Mode Transitions via Spatiotemporally-Modulated Acoustic Metamaterials

In lossless acoustic systems, mode transitions are always time-reversible, consistent with Lorentz reciprocity, giving rise to symmetric sound manipulation in space-time. To overcome this fundamental limitation and break space-time symmetry, nonreciprocal sound steering is realized by designing and experimentally implementing spatiotemporally-modulated acoustic metamaterials. Relying on no slow mechanical parts, unstable and noisy airflow or complicated piezoelectric array, our mechanism uses the coupling between an ultrathin membrane and external electromagnetic field to realize programmable, dynamic control of acoustic impedance in a motionless and noiseless manner. The fast and flexible impedance modulation at the deeply subwavelength scale enabled by our compact metamaterials provides an effective unidirectional momentum in space-time to realize irreversible transition in k-ω space between different diffraction modes. The nonreciprocal wave-steering functionality of the proposed metamaterial is elucidated by theoretically deriving the time-varying acoustic response and demonstrated both numerically and experimentally via two distinctive examples of unidirectional evanescent wave conversion and nonreciprocal blue-shift focusing. This work can be further extended into the paradigm of Bloch waves and impact other vibrant domains, such as non-Hermitian topological acoustics and parity-time-symmetric acoustics.

physics.app-ph↗

arXiv2021

Any-Precision Deep Neural Networks

We present any-precision deep neural networks (DNNs), which are trained with a new method that allows the learned DNNs to be flexible in numerical precision during inference. The same model in runtime can be flexibly and directly set to different bit-widths, by truncating the least significant bits, to support dynamic speed and accuracy trade-off. When all layers are set to low-bits, we show that the model achieved accuracy comparable to dedicated models trained at the same precision. This nice property facilitates flexible deployment of deep learning models in real-world applications, where in practice trade-offs between model accuracy and runtime efficiency are often sought. Previous literature presents solutions to train models at each individual fixed efficiency/accuracy trade-off point. But how to produce a model flexible in runtime precision is largely unexplored. When the demand of efficiency/accuracy trade-off varies from time to time or even dynamically changes in runtime, it is infeasible to re-train models accordingly, and the storage budget may forbid keeping multiple models. Our proposed framework achieves this flexibility without performance degradation. More importantly, we demonstrate that this achievement is agnostic to model architectures and applicable to multiple vision tasks. Our code is released at https://github.com/SHI-Labs/Any-Precision-DNNs.

cs.LG↗

arXiv2020

Observation of a Chiral Wave Function in Twofold Degenerate Quadruple Weyl System BaPtGe

Topological states in quantum materials are defined by non-trivial topological invariants, such as the Chern number, which are properties of their bulk wave functions. A remarkable consequence of topological wave functions is the emergence of edge modes, a phenomenon known as bulk-edge correspondence, that gives rise to quantized or chiral physical properties. While edge modes are widely presented as signatures of non-trivial topology, how bulk wave functions can manifest explicitly topological properties remains unresolved. Here, using high-resolution inelastic x-ray spectroscopy (IXS) combined with first principles calculations, we report experimental signatures of chiral wave functions in the bulk phonon spectrum of BaPtGe, which we show to host a previously undiscovered twofold degenerate quadruple Weyl node. The chirality of the degenerate phononic wave function yields a non-trivial phonon dynamical structure factor, S(Q,$ω$), along high-symmetry directions, that is in excellent agreement with numerical and model calculations. Our results establish IXS as a powerful tool to uncover topological wave functions, providing a key missing ingredient in the study of topological quantum matter.

cond-mat.mtrl-sci↗

arXiv2020

Strong spin-dephasing in a topological insulator - paramagnet heterostructure

The interface between magnetic materials and topological insulators can drive the formation of exotic phases of matter and enable functionality through manipulation of the strong spin polarized transport. Here, we report that the spin-momentum-locked transport in the topological insulator Bi$_2$Se$_3$ is completely suppressed by scattering at a heterointerface with the kagome-lattice paramagnet, Co$_7$Se$_8$. Bi$_2$Se$_{3-}$Co$_7$Se$_{8-}$Bi$_2$Se$_3$ trilayer heterostructures were grown using molecular beam epitaxy. Magnetotransport measurements revealed a substantial suppression of the weak antilocalization effect for Co$_7$Se$_8$ at thicknesses as thin as a monolayer, indicating a strong dephasing mechanism. Bi$_{2-x}$Co$_x$Se$_3$ films, where Co is in a non-magnetic $3^+$ state, show weak antilocalization that survives to $x = 0.5$, which, in comparison with the heterostructures, suggests the unordered moments of the Co$^{2+}$ act as a far stronger dephasing element. This work highlights several important points regarding spin-polarized transport in topological insulator interfaces and how magnetic materials can be integrated with topological materials to realize both exotic phases as well as novel device functionality.

cond-mat.mtrl-sci↗

arXiv2020

Few-Shot Open-Set Recognition using Meta-Learning

The problem of open-set recognition is considered. While previous approaches only consider this problem in the context of large-scale classifier training, we seek a unified solution for this and the low-shot classification setting. It is argued that the classic softmax classifier is a poor solution for open-set recognition, since it tends to overfit on the training classes. Randomization is then proposed as a solution to this problem. This suggests the use of meta-learning techniques, commonly used for few-shot classification, for the solution of open-set recognition. A new oPen sEt mEta LEaRning (PEELER) algorithm is then introduced. This combines the random selection of a set of novel classes per episode, a loss that maximizes the posterior entropy for examples of those classes, and a new metric learning formulation based on the Mahalanobis distance. Experimental results show that PEELER achieves state of the art open set recognition performance for both few-shot and large-scale recognition. On CIFAR and miniImageNet, it achieves substantial gains in seen/unseen class detection AUROC for a given seen-class classification accuracy.

cs.CV↗

arXiv2019

LeRoP: A Learning-Based Modular Robot Photography Framework

We introduce a novel framework for automatic capturing of human portraits. The framework allows the robot to follow a person to the desired location using a Person Re-identification model. When composing is activated, the robot attempts to adjust its position to form the view that can best match the given template image, and finally takes a photograph. A template image can be predicted dynamically using an off-the-shelf photo evaluation model by the framework, or selected manually from a pre-defined set by the user. The template matching-based view adjustment is driven by a deep reinforcement learning network. Our framework lies on top of the Robot Operating System (ROS). The framework is designed to be modular so that all the models can be flexibly replaced based on needs. We show our framework on a variety of examples. In particular, we tested it in three indoor scenes and used it to take 20 photos of each scene: ten for the pre-defined template, ten for the dynamically generated ones. The average number of adjustment was $11.20$ for pre-defined templates and $12.76$ for dynamically generated ones; the average time spent was $22.11$ and $24.10$ seconds respectively.

cs.RO↗

arXiv2019

Strongly Gapped Topological Surface States on Protected Surfaces of Antiferromagnetic MnBi$_4$Te$_7$ and MnBi$_6$Te$_{10}$

The search for materials to support the Quantum Anomalous Hall Effect (QAHE) have recently centered on intrinsic magnetic topological insulators (MTIs) including MnBi$_2$Te$_4$ or heterostructures made up of MnBi$_2$Te$_4$ and Bi$_2$Te$_3$. While MnBi$_2$Te$_4$ is itself a MTI, most recent ARPES experiments indicate that the surface states on this material lack the mass gap that is expected from the magnetism-induced time-reversal symmetry breaking (TRSB), with the absence of this mass gap likely due to surface magnetic disorder. Here, utilizing small-spot ARPES scanned across the surfaces of MnBi$_4$Te$_7$ and MnBi$_6$Te$_{10}$, we show the presence of large mass gaps (~ 100 meV scale) on both of these materials when the MnBi$_2$Te$_4$ surfaces are buried below one layer of Bi$_2$Te$_3$ that apparently protects the magnetic order, but not when the MnBi$_2$Te$_4$ surfaces are exposed at the surface or are buried below two Bi$_2$Te$_3$ layers. This makes both MnBi$_4$Te$_7$ and MnBi$_6$Te$_{10}$ excellent candidates for supporting the QAHE, especially if bulk devices can be fabricated with a single continuous Bi$_2$Te$_3$ layer at the surface.

cond-mat.str-el↗

arXiv2019

Towards Interpretable Face Recognition

Deep CNNs have been pushing the frontier of visual recognition over past years. Besides recognition accuracy, strong demands in understanding deep CNNs in the research community motivate developments of tools to dissect pre-trained models to visualize how they make predictions. Recent works further push the interpretability in the network learning stage to learn more meaningful representations. In this work, focusing on a specific area of visual recognition, we report our efforts towards interpretable face recognition. We propose a spatial activation diversity loss to learn more structured face representations. By leveraging the structure, we further design a feature activation diversity loss to push the interpretable representations to be discriminative and robust to occlusions. We demonstrate on three face recognition benchmarks that our proposed method is able to improve face recognition accuracy with easily interpretable face representations.

cs.CV↗

arXiv2019

Four-legged starfish-shaped Cooper pairs with ultrashort antinodal length scales in cuprate superconductors

Cooper pairs of mutually attracting electrons form the building blocks of superconductivity. Thirty years after the discovery of high-temperature superconductivity in cuprates, many details of the pairs remain unknown, including their size and shape. Here we apply brand new ARPES-based methods that allow us to reconstruct the shape and size of the pairs in Bi$_2$Sr$_2$CaCu$_2$O$_{8+δ}$. The pairs are seen to form a characteristic starfish shape that is very long (>50Å) in the near-nodal direction but extremely short (~4.5Å) in the antinodal (Cu-O) direction. We find that this ultrashort antinodal length scale, which is of order a lattice constant, is approximately constant over a wide range of doping levels even as many other parameters including the pairing strength change. This suggests that this new length scale, along with the pair shape, is one of the most fundamental characteristics of the pairs. Further, the shape and ultrashort length scale should make the pairs create or intertwine with variations in charge and pair density, center on various types of lattice positions, and potentially explain aspects of the nematic order in these materials.

cond-mat.supr-con↗

arXiv2019

Towards Physically Safe Reinforcement Learning under Supervision

This paper addresses the question of how a previously available control policy $π_s$ can be used as a supervisor to more quickly and safely train a new learned control policy $π_L$ for a robot. A weighted average of the supervisor and learned policies is used during trials, with a heavier weight initially on the supervisor, in order to allow safe and useful physical trials while the learned policy is still ineffective. During the process, the weight is adjusted to favor the learned policy. As weights are adjusted, the learned network must compensate so as to give safe and reasonable outputs under the different weights. A pioneer network is introduced that pre-learns a policy that performs similarly to the current learned policy under the planned next step for new weights; this pioneer network then replaces the currently learned network in the next set of trials. Experiments in OpenAI Gym demonstrate the effectiveness of the proposed method.

cs.LG↗

arXiv2018

A Modulation Module for Multi-task Learning with Applications in Image Retrieval

Multi-task learning has been widely adopted in many computer vision tasks to improve overall computation efficiency or boost the performance of individual tasks, under the assumption that those tasks are correlated and complementary to each other. However, the relationships between the tasks are complicated in practice, especially when the number of involved tasks scales up. When two tasks are of weak relevance, they may compete or even distract each other during joint training of shared parameters, and as a consequence undermine the learning of all the tasks. This will raise destructive interference which decreases learning efficiency of shared parameters and lead to low quality loss local optimum w.r.t. shared parameters. To address the this problem, we propose a general modulation module, which can be inserted into any convolutional neural network architecture, to encourage the coupling and feature sharing of relevant tasks while disentangling the learning of irrelevant tasks with minor parameters addition. Equipped with this module, gradient directions from different tasks can be enforced to be consistent for those shared parameters, which benefits multi-task joint training. The module is end-to-end learnable without ad-hoc design for specific tasks, and can naturally handle many tasks at the same time. We apply our approach on two retrieval tasks, face retrieval on the CelebA dataset [1] and product retrieval on the UT-Zappos50K dataset [2, 3], and demonstrate its advantage over other multi-task learning methods in both accuracy and storage efficiency.

cs.CV↗

arXiv2018

Contemplating Visual Emotions: Understanding and Overcoming Dataset Bias

While machine learning approaches to visual emotion recognition offer great promise, current methods consider training and testing models on small scale datasets covering limited visual emotion concepts. Our analysis identifies an important but long overlooked issue of existing visual emotion benchmarks in the form of dataset biases. We design a series of tests to show and measure how such dataset biases obstruct learning a generalizable emotion recognition model. Based on our analysis, we propose a webly supervised approach by leveraging a large quantity of stock image data. Our approach uses a simple yet effective curriculum guided training strategy for learning discriminative emotion features. We discover that the models learned using our large scale stock image dataset exhibit significantly better generalization ability than the existing datasets without the manual collection of even a single label. Moreover, visual representation learned using our approach holds a lot of promise across a variety of tasks on different image and video datasets.

cs.CV↗

arXiv2018

Active Object Perceiver: Recognition-guided Policy Learning for Object Searching on Mobile Robots

We study the problem of learning a navigation policy for a robot to actively search for an object of interest in an indoor environment solely from its visual inputs. While scene-driven visual navigation has been widely studied, prior efforts on learning navigation policies for robots to find objects are limited. The problem is often more challenging than target scene finding as the target objects can be very small in the view and can be in an arbitrary pose. We approach the problem from an active perceiver perspective, and propose a novel framework that integrates a deep neural network based object recognition module and a deep reinforcement learning based action prediction mechanism. To validate our method, we conduct experiments on both a simulation dataset (AI2-THOR) and a real-world environment with a physical robot. We further propose a new decaying reward function to learn the control policy specific to the object searching task. Experimental results validate the efficacy of our method, which outperforms competing methods in both average trajectory length and success rate.

cs.AI↗

arXiv2018

Observation of Topological Surface State in High Temperature Superconductor MgB2

The hunt for the benchmark topological superconductor (TSc) has been an extremely active research subject in condensed matter research, with quite a few candidates identified or proposed. However, low transition temperatures (Tc) and/or strong sensitivity to disorder and dopant levels in known TSc candidates have greatly hampered progress in this field. Here, we use Angle-resolved Photoemission Spectroscopy (ARPES) to show the presence of Dirac Nodal Lines (DNLs) and the corresponding topological surface states (TSS's) on the [010] faces of the Tc=39K s-wave BCS superconductor MgB2. Not only is this nearly triple the current record of superconducting Tc among all candidate TSc's, but the nature of these DNL states should make them highly tolerant against disorder and inadvertent doping variations. This makes MgB2 a promising high temperature platform for the study of topological superconductivity.

cond-mat.supr-con↗

arXiv2018

Spectroscopic Evidence of Low Energy Gaps Persisting Towards 120 Kelvin in Surface-Doped p-Terphenyl Crystals

The possibility of high temperature superconductivity in organic compounds has been discussed since the pioneering work of Little in 1964, with unsatisfactory progress until the recent report of a weak Meissner shielding effect at 120 Kelvin in potassium-doped para-terphenyl samples. To date however, no other signals of the superconductivity have been shown, including the zero-resistance state or evidence for the formation of the Cooper pairs that are inherent to the superconducting state. Here, using high-resolution photoemission spectroscopy on potassium surface-doped para-terphenyl crystals, we uncover low energy gaps that persist to approximately 120 K. Among a few potential origins for these gaps, we argue that the onset of electron pairing within molecules is most likely. And while pairing gaps are a prerequisite for high temperature superconductivity they do not guarantee it. Rather, the development of long-range phase coherence between the paired states on the molecules is necessary, requiring good wavefunction overlap between molecular states--something that is in general difficult for such weakly overlapping molecules.

cond-mat.supr-con↗

arXiv2018

The AdobeIndoorNav Dataset: Towards Deep Reinforcement Learning based Real-world Indoor Robot Visual Navigation

Deep reinforcement learning (DRL) demonstrates its potential in learning a model-free navigation policy for robot visual navigation. However, the data-demanding algorithm relies on a large number of navigation trajectories in training. Existing datasets supporting training such robot navigation algorithms consist of either 3D synthetic scenes or reconstructed scenes. Synthetic data suffers from domain gap to the real-world scenes while visual inputs rendered from 3D reconstructed scenes have undesired holes and artifacts. In this paper, we present a new dataset collected in real-world to facilitate the research in DRL based visual navigation. Our dataset includes 3D reconstruction for real-world scenes as well as densely captured real 2D images from the scenes. It provides high-quality visual inputs with real-world scene complexity to the robot at dense grid locations. We further study and benchmark one recent DRL based navigation algorithm and present our attempts and thoughts on improving its generalizability to unseen test targets in the scenes.

cs.RO↗

arXiv2017

Coexistence of Tunable Weyl Points and Topological Nodal Lines in Ternary Transition-Metal Telluride TaIrTe4

We report a combined theoretical and experimental study on TaIrTe4, a potential candidate of the minimal model of type-II Weyl semimetals. Unexpectedly, an intriguing node structure with twelve Weyl points and a pair of nodal lines protected by mirror symmetry was found by first-principle calculations, with its complex signatures such as the topologically non-trivial band crossings and topologically trivial Fermi arcs cross-validated by angle-resolved photoemission spectroscopy. Through external strain, the number of Weyl points can be reduced to the theoretical minimum of four, and the appearance of the nodal lines can be switched between different mirror planes in momentum space. The coexistence of tunable Weyl points and nodal lines establishes ternary transition-metal tellurides as a unique test ground for topological state characterization and engineering.

cond-mat.mtrl-sci↗

arXiv2017

Learning Dense Facial Correspondences in Unconstrained Images

We present a minimalistic but effective neural network that computes dense facial correspondences in highly unconstrained RGB images. Our network learns a per-pixel flow and a matchability mask between 2D input photographs of a person and the projection of a textured 3D face model. To train such a network, we generate a massive dataset of synthetic faces with dense labels using renderings of a morphable face model with variations in pose, expressions, lighting, and occlusions. We found that a training refinement using real photographs is required to drastically improve the ability to handle real images. When combined with a facial detection and 3D face fitting step, we show that our approach outperforms the state-of-the-art face alignment methods in terms of accuracy and speed. By directly estimating dense correspondences, we do not rely on the full visibility of sparse facial landmarks and are not limited to the model space of regression-based approaches. We also assess our method on video frames and demonstrate successful per-frame processing under extreme pose variations, occlusions, and lighting conditions. Compared to existing 3D facial tracking techniques, our fitting does not rely on previous frames or frontal facial initialization and is robust to imperfect face detections.

cs.CV↗