arXiv subjects

Bo Han

Publications and source records attributed to Bo Han.

At least 325 records · Page 18Linked to original sources

arXiv2021

Valley-exchange coupling probed by angle-resolved photoluminescence

The optical properties of monolayer transition metal dichalcogenides are dominated by tightly-bound excitons. They form at distinct valleys in reciprocal space, and can interact via the valley-exchange coupling, modifying their dispersion considerably. Here, we predict that angle-resolved photoluminescence can be used to probe the changes of the excitonic dispersion. The exchange-coupling leads to a unique angle dependence of the emission intensity for both circularly and linearly-polarised light. We show that these emission characteristics can be strongly tuned by an external magnetic field due to the valley-specific Zeeman-shift. We propose that angle-dependent photoluminescence measurements involving both circular and linear optical polarisation as well as magnetic fields should act as strong verification of the role of valley-exchange coupling on excitonic dispersionand its signatures in optical spectra

cond-mat.mes-hall↗

arXiv2021

Spatial coherence of room-temperature monolayer WSe$_2$ exciton-polaritons in a trap

The emergence of spatial and temporal coherence of light emitted from solid-state systems is a fundamental phenomenon, rooting in a plethora of microscopic processes. It is intrinsically aligned with the control of light-matter coupling, and canonical for laser oscillation. However, it also emerges in the superradiance of multiple, phase-locked emitters, and more recently, coherence and long-range order have been investigated in bosonic condensates of thermalized light, as well as in exciton-polaritons driven to a ground state via stimulated scattering. Here, we experimentally show that the interaction between photons in a Fabry-Perot microcavity and excitons in an atomically thin WSe$_2$ layer is sufficient such that the system enters the hybridized regime of strong light-matter coupling at ambient conditions. Via Michelson interferometry, we capture clear evidence of increased spatial and temporal coherence of the emitted light from the spatially confined system ground-state. The coherence build-up is accompanied by a threshold-like behaviour of the emitted light intensity, which is a fingerprint of a polariton laser effect. Valley-physics is manifested in the presence of an external magnetic field, which allows us to manipulate K and K' polaritons via the Valley-Zeeman-effect. Our findings are of high application relevance, as they confirm the possibility to use atomically thin crystals as simple and versatile components of coherent light-sources, and in valleytronic applications at room temperature.

cond-mat.mes-hall↗

arXiv2021

Provably End-to-end Label-Noise Learning without Anchor Points

In label-noise learning, the transition matrix plays a key role in building statistically consistent classifiers. Existing consistent estimators for the transition matrix have been developed by exploiting anchor points. However, the anchor-point assumption is not always satisfied in real scenarios. In this paper, we propose an end-to-end framework for solving label-noise learning without anchor points, in which we simultaneously optimize two objectives: the cross entropy loss between the noisy label and the predicted probability by the neural network, and the volume of the simplex formed by the columns of the transition matrix. Our proposed framework can identify the transition matrix if the clean class-posterior probabilities are sufficiently scattered. This is by far the mildest assumption under which the transition matrix is provably identifiable and the learned classifier is statistically consistent. Experimental results on benchmark datasets demonstrate the effectiveness and robustness of the proposed method.

cs.LG↗

arXiv2021

Torsion, energy magnetization and thermal Hall effect

We study the effective action of hydrostatic response to torsion in the absence of spin connections in gapped $\left(2+1\right)$-dimensional topological phases. In previous studies, a torsional Chern-Simons term with a temperature-squared ($T^2$) coefficient was proposed as an alternative action to describe thermal Hall effect with the idea of balancing the diffusion of heat by a torsional field. However, the question remains whether this action leads to local bulk thermal response which is not suppressed by the gap. In our hydrostatic effective action, we show that the $T^2$ bulk term is invariant under variations up to boundary terms considering the back reaction of the geometry on local temperature, which precisely describes the edge thermal current. Furthermore, there is no boundary diffeomorphism anomalies and bulk inflow thermal currents at equilibrium and therefore no edge-to-edge adiabatic thermal current pumping. These results are in consistent with exponentially suppressed thermal current for gapped phases.

cond-mat.mes-hall↗

arXiv2021

Click-through Rate Prediction with Auto-Quantized Contrastive Learning

Click-through rate (CTR) prediction becomes indispensable in ubiquitous web recommendation applications. Nevertheless, the current methods are struggling under the cold-start scenarios where the user interactions are extremely sparse. We consider this problem as an automatic identification about whether the user behaviors are rich enough to capture the interests for prediction, and propose an Auto-Quantized Contrastive Learning (AQCL) loss to regularize the model. Different from previous methods, AQCL explores both the instance-instance and the instance-cluster similarity to robustify the latent representation, and automatically reduces the information loss to the active users due to the quantization. The proposed framework is agnostic to different model architectures and can be trained in an end-to-end fashion. Extensive results show that it consistently improves the current state-of-the-art CTR models.

cs.IR↗

arXiv2021

MC$^2$-SF: Slow-Fast Learning for Mobile-Cloud Collaborative Recommendation

With the hardware development of mobile devices, it is possible to build the recommendation models on the mobile side to utilize the fine-grained features and the real-time feedbacks. Compared to the straightforward mobile-based modeling appended to the cloud-based modeling, we propose a Slow-Fast learning mechanism to make the Mobile-Cloud Collaborative recommendation (MC$^2$-SF) mutual benefit. Specially, in our MC$^2$-SF, the cloud-based model and the mobile-based model are respectively treated as the slow component and the fast component, according to their interaction frequency in real-world scenarios. During training and serving, they will communicate the prior/privileged knowledge to each other to help better capture the user interests about the candidates, resembling the role of System I and System II in the human cognition. We conduct the extensive experiments on three benchmark datasets and demonstrate the proposed MC$^2$-SF outperforms several state-of-the-art methods.

cs.IR↗

arXiv2021

Engineering of Atomic-Scale Flexoelectricity at Grain Boundaries

Flexoelectricity is a type of ubiquitous and prominent electromechanical coupling, pertaining to the response of electrical polarization to mechanical strain gradients while not restricted to the symmetry of materials. However, large elastic deformation in most solids is usually difficult to achieve and the strain gradient at minuscule is challenging to control. Here we exploit the exotic structural inhomogeneity of grain boundary to achieve a huge strain gradient (~ 1.2 nm-1) within 3 ~ 4 unit-cells, and thus obtain atomic-scale flexoelectric polarization up to ~ 38 μC/cm2 at a 24 LaAlO3 grain boundary. The nanoscale flexoelectricity also modifies the electrical activity of grain boundaries. Moreover, we prove that it is a general and feasible way to form large strain gradients at atomic scale by altering the misorientation angles of grain boundaries in different dielectric materials. Thus, engineering of grain boundaries provides an effective pathway to achieve tunable flexoelectricity and broadens the electromechanical functionalities of non-piezoelectric materials.

cond-mat.mtrl-sci↗

arXiv2021

Maximum Mean Discrepancy Test is Aware of Adversarial Attacks

The maximum mean discrepancy (MMD) test could in principle detect any distributional discrepancy between two datasets. However, it has been shown that the MMD test is unaware of adversarial attacks -- the MMD test failed to detect the discrepancy between natural and adversarial data. Given this phenomenon, we raise a question: are natural and adversarial data really from different distributions? The answer is affirmative -- the previous use of the MMD test on the purpose missed three key factors, and accordingly, we propose three components. Firstly, the Gaussian kernel has limited representation power, and we replace it with an effective deep kernel. Secondly, the test power of the MMD test was neglected, and we maximize it following asymptotic statistics. Finally, adversarial data may be non-independent, and we overcome this issue with the wild bootstrap. By taking care of the three factors, we verify that the MMD test is aware of adversarial attacks, which lights up a novel road for adversarial data detection based on two-sample tests.

cs.LG↗

arXiv2021

Local Reweighting for Adversarial Training

Instances-reweighted adversarial training (IRAT) can significantly boost the robustness of trained models, where data being less/more vulnerable to the given attack are assigned smaller/larger weights during training. However, when tested on attacks different from the given attack simulated in training, the robustness may drop significantly (e.g., even worse than no reweighting). In this paper, we study this problem and propose our solution--locally reweighted adversarial training (LRAT). The rationale behind IRAT is that we do not need to pay much attention to an instance that is already safe under the attack. We argue that the safeness should be attack-dependent, so that for the same instance, its weight can change given different attacks based on the same model. Thus, if the attack simulated in training is mis-specified, the weights of IRAT are misleading. To this end, LRAT pairs each instance with its adversarial variants and performs local reweighting inside each pair, while performing no global reweighting--the rationale is to fit the instance itself if it is immune to the attack, but not to skip the pair, in order to passively defend different attacks in future. Experiments show that LRAT works better than both IRAT (i.e., global reweighting) and the standard AT (i.e., no reweighting) when trained with an attack and tested on different attacks.

cs.LG↗

arXiv2021

Dual T: Reducing Estimation Error for Transition Matrix in Label-noise Learning

The transition matrix, denoting the transition relationship from clean labels to noisy labels, is essential to build statistically consistent classifiers in label-noise learning. Existing methods for estimating the transition matrix rely heavily on estimating the noisy class posterior. However, the estimation error for noisy class posterior could be large due to the randomness of label noise, which would lead the transition matrix to be poorly estimated. Therefore, in this paper, we aim to solve this problem by exploiting the divide-and-conquer paradigm. Specifically, we introduce an intermediate class to avoid directly estimating the noisy class posterior. By this intermediate class, the original transition matrix can then be factorized into the product of two easy-to-estimate transition matrices. We term the proposed method the dual-T estimator. Both theoretical analyses and empirical results illustrate the effectiveness of the dual-T estimator for estimating transition matrices, leading to better classification performances.

cs.LG↗

arXiv2021

Atomic-scale imaging of CH3NH3PbI3 structure and its decomposition pathway

Understanding the atomic structure and structural instability of organic-inorganic hybrid perovskites is the key to appreciate their remarkable photoelectric properties and failure mechanism. Here, using low-dose imaging technique by direct-detection electron-counting camera in transmission electron microscope, we investigate the atomic structure and decomposition pathway of CH3NH3PbI3 (MAPbI3) at the atomic scale. We successfully image the atomic structure of perovskite in real space under ultra-low electron dose condition, and observe a two-step decomposition process, i.e. initial loss of MA followed by the collapse of perovskite structure into 6H-PbI2 with their critical threshold dose also determined. Interestingly, an intermediate phase (MA0.5PbI3) with locally ordered vacancies can robustly exist before perovskite collapses, enlightening strategies for prevention and recovery of perovskite structure during degradation. Associated with structure evolution, the bandgap gradually increases from ~1.6 eV to ~2.1 eV, and it is found that both C-N and N-H bonds can be destroyed under irradiation, releasing NH3 and leaving hydrocarbons. These findings enhance our understanding of the photoelectric properties and failure mechanism of MAPbI3, providing potential strategy into material optimization.

physics.app-ph↗

arXiv2021

Learning Diverse-Structured Networks for Adversarial Robustness

In adversarial training (AT), the main focus has been the objective and optimizer while the model has been less studied, so that the models being used are still those classic ones in standard training (ST). Classic network architectures (NAs) are generally worse than searched NAs in ST, which should be the same in AT. In this paper, we argue that NA and AT cannot be handled independently, since given a dataset, the optimal NA in ST would be no longer optimal in AT. That being said, AT is time-consuming itself; if we directly search NAs in AT over large search spaces, the computation will be practically infeasible. Thus, we propose a diverse-structured network (DS-Net), to significantly reduce the size of the search space: instead of low-level operations, we only consider predefined atomic blocks, where an atomic block is a time-tested building block like the residual block. There are only a few atomic blocks and thus we can weight all atomic blocks rather than find the best one in a searched block of DS-Net, which is an essential trade-off between exploring diverse structures and exploiting the best structures. Empirical results demonstrate the advantages of DS-Net, i.e., weighting the atomic blocks.

cs.LG↗

arXiv2021

Class2Simi: A Noise Reduction Perspective on Learning with Noisy Labels

Learning with noisy labels has attracted a lot of attention in recent years, where the mainstream approaches are in pointwise manners. Meanwhile, pairwise manners have shown great potential in supervised metric learning and unsupervised contrastive learning. Thus, a natural question is raised: does learning in a pairwise manner mitigate label noise? To give an affirmative answer, in this paper, we propose a framework called Class2Simi: it transforms data points with noisy class labels to data pairs with noisy similarity labels, where a similarity label denotes whether a pair shares the class label or not. Through this transformation, the reduction of the noise rate is theoretically guaranteed, and hence it is in principle easier to handle noisy similarity labels. Amazingly, DNNs that predict the clean class labels can be trained from noisy data pairs if they are first pretrained from noisy data points. Class2Simi is computationally efficient because not only this transformation is on-the-fly in mini-batches, but also it just changes loss computation on top of model prediction into a pairwise manner. Its effectiveness is verified by extensive experiments.

cs.LG↗

arXiv2021

Device-Cloud Collaborative Learning for Recommendation

With the rapid development of storage and computing power on mobile devices, it becomes critical and popular to deploy models on devices to save onerous communication latencies and to capture real-time features. While quite a lot of works have explored to facilitate on-device learning and inference, most of them focus on dealing with response delay or privacy protection. Little has been done to model the collaboration between the device and the cloud modeling and benefit both sides jointly. To bridge this gap, we are among the first attempts to study the Device-Cloud Collaborative Learning (DCCL) framework. Specifically, we propose a novel MetaPatch learning approach on the device side to efficiently achieve "thousands of people with thousands of models" given a centralized cloud model. Then, with billions of updated personalized device models, we propose a "model-over-models" distillation algorithm, namely MoMoDistill, to update the centralized cloud model. Our extensive experiments over a range of datasets with different settings demonstrate the effectiveness of such collaboration on both cloud and devices, especially its superiority to model long-tailed users.

cs.LG↗

arXiv2021

Towards Defending against Adversarial Examples via Attack-Invariant Features

Deep neural networks (DNNs) are vulnerable to adversarial noise. Their adversarial robustness can be improved by exploiting adversarial examples. However, given the continuously evolving attacks, models trained on seen types of adversarial examples generally cannot generalize well to unseen types of adversarial examples. To solve this problem, in this paper, we propose to remove adversarial noise by learning generalizable invariant features across attacks which maintain semantic classification information. Specifically, we introduce an adversarial feature learning mechanism to disentangle invariant features from adversarial noise. A normalization term has been proposed in the encoded space of the attack-invariant features to address the bias issue between the seen and unseen types of attacks. Empirical evaluations demonstrate that our method could provide better protection in comparison to previous state-of-the-art approaches, especially against unseen types of attacks and adaptive attacks.

cs.CV↗

arXiv2021

Sample Selection with Uncertainty of Losses for Learning with Noisy Labels

In learning with noisy labels, the sample selection approach is very popular, which regards small-loss data as correctly labeled during training. However, losses are generated on-the-fly based on the model being trained with noisy labels, and thus large-loss data are likely but not certainly to be incorrect. There are actually two possibilities of a large-loss data point: (a) it is mislabeled, and then its loss decreases slower than other data, since deep neural networks "learn patterns first"; (b) it belongs to an underrepresented group of data and has not been selected yet. In this paper, we incorporate the uncertainty of losses by adopting interval estimation instead of point estimation of losses, where lower bounds of the confidence intervals of losses derived from distribution-free concentration inequalities, but not losses themselves, are used for sample selection. In this way, we also give large-loss but less selected data a try; then, we can better distinguish between the cases (a) and (b) by seeing if the losses effectively decrease with the uncertainty after the try. As a result, we can better explore underrepresented data that are correctly labeled but seem to be mislabeled at first glance. Experiments demonstrate that the proposed method is superior to baselines and robust to a broad range of label noise types.

cs.LG↗

arXiv2021

Instance Correction for Learning with Open-set Noisy Labels

The problem of open-set noisy labels denotes that part of training data have a different label space that does not contain the true class. Lots of approaches, e.g., loss correction and label correction, cannot handle such open-set noisy labels well, since they need training data and test data to share the same label space, which does not hold for learning with open-set noisy labels. The state-of-the-art methods thus employ the sample selection approach to handle open-set noisy labels, which tries to select clean data from noisy data for network parameters updates. The discarded data are seen to be mislabeled and do not participate in training. Such an approach is intuitive and reasonable at first glance. However, a natural question could be raised "can such data only be discarded during training?". In this paper, we show that the answer is no. Specifically, we discuss that the instances of discarded data could consist of some meaningful information for generalization. For this reason, we do not abandon such data, but use instance correction to modify the instances of the discarded data, which makes the predictions for the discarded data consistent with given labels. Instance correction are performed by targeted adversarial attacks. The corrected data are then exploited for training to help generalization. In addition to the analytical results, a series of empirical evidences are provided to justify our claims.

cs.LG↗

arXiv2021

Geometry-aware Instance-reweighted Adversarial Training

In adversarial machine learning, there was a common belief that robustness and accuracy hurt each other. The belief was challenged by recent studies where we can maintain the robustness and improve the accuracy. However, the other direction, whether we can keep the accuracy while improving the robustness, is conceptually and practically more interesting, since robust accuracy should be lower than standard accuracy for any model. In this paper, we show this direction is also promising. Firstly, we find even over-parameterized deep networks may still have insufficient model capacity, because adversarial training has an overwhelming smoothing effect. Secondly, given limited model capacity, we argue adversarial data should have unequal importance: geometrically speaking, a natural data point closer to/farther from the class boundary is less/more robust, and the corresponding adversarial data point should be assigned with larger/smaller weight. Finally, to implement the idea, we propose geometry-aware instance-reweighted adversarial training, where the weights are based on how difficult it is to attack a natural data point. Experiments show that our proposal boosts the robustness of standard adversarial training; combining two directions, we improve both robustness and accuracy of standard adversarial training.

cs.LG↗