arXiv subjects

Tao Yu

Publications and source records attributed to Tao Yu.

At least 271 records · Page 15Linked to original sources

arXiv2022

Efficient Gating of Magnons by Proximity Superconductors

Electrostatic gating confines and controls the transport of electrons in integrated circuits. Magnons, the quanta of spin waves of the magnetic order, are promising alternative information carriers, but difficult to gate. Here we report that superconducting strips on top of thin magnetic films can totally reflect magnons by its diamagnetic response to the magnon stray fields. The induced large frequency shifts unidirectionally blocks the magnons propagating normal to the magnetization. Two superconducting gates parallel to the magnetization create a magnonic cavity. The option to gate coherent magnons adds functionalities to magnonic devices, such as reprogrammable logical devices and increased couplings to other degrees of freedom.

cond-mat.mes-hall↗

arXiv2022

GIMO: Gaze-Informed Human Motion Prediction in Context

Predicting human motion is critical for assistive robots and AR/VR applications, where the interaction with humans needs to be safe and comfortable. Meanwhile, an accurate prediction depends on understanding both the scene context and human intentions. Even though many works study scene-aware human motion prediction, the latter is largely underexplored due to the lack of ego-centric views that disclose human intent and the limited diversity in motion and scenes. To reduce the gap, we propose a large-scale human motion dataset that delivers high-quality body pose sequences, scene scans, as well as ego-centric views with the eye gaze that serves as a surrogate for inferring human intent. By employing inertial sensors for motion capture, our data collection is not tied to specific scenes, which further boosts the motion dynamics observed from our subjects. We perform an extensive study of the benefits of leveraging the eye gaze for ego-centric human motion prediction with various state-of-the-art architectures. Moreover, to realize the full potential of the gaze, we propose a novel network architecture that enables bidirectional communication between the gaze and motion branches. Our network achieves the top performance in human motion prediction on the proposed dataset, thanks to the intent information from eye gaze and the denoised gaze feature modulated by the motion. Code and data can be found at https://github.com/y-zheng18/GIMO.

cs.CV↗

arXiv2022

Geometry-aware Single-image Full-body Human Relighting

Single-image human relighting aims to relight a target human under new lighting conditions by decomposing the input image into albedo, shape and lighting. Although plausible relighting results can be achieved, previous methods suffer from both the entanglement between albedo and lighting and the lack of hard shadows, which significantly decrease the realism. To tackle these two problems, we propose a geometry-aware single-image human relighting framework that leverages single-image geometry reconstruction for joint deployment of traditional graphics rendering and neural rendering techniques. For the de-lighting, we explore the shortcomings of UNet architecture and propose a modified HRNet, achieving better disentanglement between albedo and lighting. For the relighting, we introduce a ray tracing-based per-pixel lighting representation that explicitly models high-frequency shadows and propose a learning-based shading refinement module to restore realistic shadows (including hard cast shadows) from the ray-traced shading maps. Our framework is able to generate photo-realistic high-frequency shadows such as cast shadows under challenging lighting conditions. Extensive experiments demonstrate that our proposed method outperforms previous methods on both synthetic and real images.

cs.CV↗

arXiv2022

Asymptotic Uncertainty of False Discovery Proportion

Multiple testing has been a popular topic in statistical research. Although vast works have been done, controlling the false discoveries remains a challenging task when the corresponding test statistics are dependent. Various methods have been proposed to estimate the false discovery proportion (FDP) under arbitrary dependence among the test statistics. One of the main ideas is to reduce arbitrary dependence to weak dependence and then to establish theoretically the strong consistency of the FDP and false discovery rate (FDR) under weak dependence. As a consequence, FDPs share the same asymptotic limit in the framework of weak dependence. We observe that the asymptotic variance of the FDP, however, may rely heavily on the dependence structure of the corresponding test statistics even when they are only weakly dependent; and it is of great practical value to quantify this variability, as it can serve as an indicator of the quality of the FDP estimate from the given data. As far as we are aware, the research on this respect is still limited in the literature. In this paper, we first derive the asymptotic expansion of FDP under mild regularity conditions and then examine how the asymptotic variance of FDP varies under different dependence structures both theoretically and numerically. With the observations in this study, we recommend that in a multiple testing performed by an FDP procedure, we may report both the mean and the variance estimates of FDP to enrich the study outcome.

math.ST↗

arXiv2022

Design and Analysis of Robust Resilient Diffusion over Multi-Task Networks Against Byzantine Attacks

This paper studies distributed diffusion adaptation over clustered multi-task networks in the presence of impulsive interferences and Byzantine attacks. We develop a robust resilient diffusion least mean Geman-McClure-estimation (RDLMG) algorithm based on the cost function used by the Geman-McClure estimator, which can reduce the sensitivity to large outliers and make the algorithm robust under impulsive interferences. Moreover, the mean sub-sequence reduced method, in which each node discards the extreme value information of cost contributions received from its neighbors, can make the network resilient against Byzantine attacks. In this regard, the proposed RDLMG algorithm ensures that all normal nodes converge to their ideal states with cooperation among nodes. A statistical analysis of the RDLMG algorithm is also carried out in terms of mean and mean-square performances. Numerical results evaluate the proposed RDLMG algorithm in applications to multi-target localization and multi-task spectrum sensing.

cs.LG↗

arXiv2022

FaceVerse: a Fine-grained and Detail-controllable 3D Face Morphable Model from a Hybrid Dataset

We present FaceVerse, a fine-grained 3D Neural Face Model, which is built from hybrid East Asian face datasets containing 60K fused RGB-D images and 2K high-fidelity 3D head scan models. A novel coarse-to-fine structure is proposed to take better advantage of our hybrid dataset. In the coarse module, we generate a base parametric model from large-scale RGB-D images, which is able to predict accurate rough 3D face models in different genders, ages, etc. Then in the fine module, a conditional StyleGAN architecture trained with high-fidelity scan models is introduced to enrich elaborate facial geometric and texture details. Note that different from previous methods, our base and detailed modules are both changeable, which enables an innovative application of adjusting both the basic attributes and the facial details of 3D face models. Furthermore, we propose a single-image fitting framework based on differentiable rendering. Rich experiments show that our method outperforms the state-of-the-art methods.

cs.CV↗

arXiv2022

SwinIQA: Learned Swin Distance for Compressed Image Quality Assessment

Image compression has raised widespread interest recently due to its significant importance for multimedia storage and transmission. Meanwhile, a reliable image quality assessment (IQA) for compressed images can not only help to verify the performance of various compression algorithms but also help to guide the compression optimization in turn. In this paper, we design a full-reference image quality assessment metric SwinIQA to measure the perceptual quality of compressed images in a learned Swin distance space. It is known that the compression artifacts are usually non-uniformly distributed with diverse distortion types and degrees. To warp the compressed images into the shared representation space while maintaining the complex distortion information, we extract the hierarchical feature representations from each stage of the Swin Transformer. Besides, we utilize cross attention operation to map the extracted feature representations into a learned Swin distance space. Experimental results show that the proposed metric achieves higher consistency with human's perceptual judgment compared with both traditional methods and learning-based methods on CLIC datasets.

cs.CV↗

arXiv2022

DeepCloth: Neural Garment Representation for Shape and Style Editing

Garment representation, editing and animation are challenging topics in the area of computer vision and graphics. It remains difficult for existing garment representations to achieve smooth and plausible transitions between different shapes and topologies. In this work, we introduce, DeepCloth, a unified framework for garment representation, reconstruction, animation and editing. Our unified framework contains 3 components: First, we represent the garment geometry with a "topology-aware UV-position map", which allows for the unified description of various garments with different shapes and topologies by introducing an additional topology-aware UV-mask for the UV-position map. Second, to further enable garment reconstruction and editing, we contribute a method to embed the UV-based representations into a continuous feature space, which enables garment shape reconstruction and editing by optimization and control in the latent space, respectively. Finally, we propose a garment animation method by unifying our neural garment representation with body shape and pose, which achieves plausible garment animation results leveraging the dynamic information encoded by our shape and style representation, even under drastic garment editing operations. To conclude, with DeepCloth, we move a step forward in establishing a more flexible and general 3D garment digitization framework. Experiments demonstrate that our method can achieve state-of-the-art garment representation performance compared with previous methods.

cs.CV↗

arXiv2022

DYLE: Dynamic Latent Extraction for Abstractive Long-Input Summarization

Transformer-based models have achieved state-of-the-art performance on short-input summarization. However, they still struggle with summarizing longer text. In this paper, we present DYLE, a novel dynamic latent extraction approach for abstractive long-input summarization. DYLE jointly trains an extractor and a generator and treats the extracted text snippets as the latent variable, allowing dynamic snippet-level attention weights during decoding. To provide adequate supervision, we propose simple yet effective heuristics for oracle extraction as well as a consistency loss term, which encourages the extractor to approximate the averaged dynamic weights predicted by the generator. We evaluate our method on different long-document and long-dialogue summarization tasks: GovReport, QMSum, and arXiv. Experiment results show that DYLE outperforms all existing methods on GovReport and QMSum, with gains up to 6.1 ROUGE, while yielding strong results on arXiv. Further analysis shows that the proposed dynamic weights provide interpretability of our generation process.

cs.CL↗

arXiv2022

ProbNVS: Fast Novel View Synthesis with Learned Probability-Guided Sampling

Existing state-of-the-art novel view synthesis methods rely on either fairly accurate 3D geometry estimation or sampling of the entire space for neural volumetric rendering, which limit the overall efficiency. In order to improve the rendering efficiency by reducing sampling points without sacrificing rendering quality, we propose to build a novel view synthesis framework based on learned MVS priors that enables general, fast and photo-realistic view synthesis simultaneously. Specifically, fewer but important points are sampled under the guidance of depth probability distributions extracted from the learned MVS architecture. Based on the learned probability-guided sampling, a neural volume rendering module is elaborately devised to fully aggregate source view information as well as the learned scene structures to synthesize photorealistic target view images. Finally, the rendering results in uncertain, occluded and unreferenced regions can be further improved by incorporating a confidence-aware refinement module. Experiments show that our method achieves 15 to 40 times faster rendering compared to state-of-the-art baselines, with strong generalization capacity and comparable high-quality novel view synthesis performance.

cs.CV↗

arXiv2022

OPAL: Occlusion Pattern Aware Loss for Unsupervised Light Field Disparity Estimation

Light field disparity estimation is an essential task in computer vision with various applications. Although supervised learning-based methods have achieved both higher accuracy and efficiency than traditional optimization-based methods, the dependency on ground-truth disparity for training limits the overall generalization performance not to say for real-world scenarios where the ground-truth disparity is hard to capture. In this paper, we argue that unsupervised methods can achieve comparable accuracy, but, more importantly, much higher generalization capacity and efficiency than supervised methods. Specifically, we present the Occlusion Pattern Aware Loss, named OPAL, which successfully extracts and encodes the general occlusion patterns inherent in the light field for loss calculation. OPAL enables: i) accurate and robust estimation by effectively handling occlusions without using any ground-truth information for training and ii) much efficient performance by significantly reducing the network parameters required for accurate inference. Besides, a transformer-based network and a refinement module are proposed for achieving even more accurate results. Extensive experiments demonstrate our method not only significantly improves the accuracy compared with the SOTA unsupervised methods, but also possesses strong generalization capacity, even for real-world data, compared with supervised methods. Our code will be made publicly available.

cs.CV↗

arXiv2022

Structured Local Radiance Fields for Human Avatar Modeling

It is extremely challenging to create an animatable clothed human avatar from RGB videos, especially for loose clothes due to the difficulties in motion modeling. To address this problem, we introduce a novel representation on the basis of recent neural scene rendering techniques. The core of our representation is a set of structured local radiance fields, which are anchored to the pre-defined nodes sampled on a statistical human body template. These local radiance fields not only leverage the flexibility of implicit representation in shape and appearance modeling, but also factorize cloth deformations into skeleton motions, node residual translations and the dynamic detail variations inside each individual radiance field. To learn our representation from RGB data and facilitate pose generalization, we propose to learn the node translations and the detail variations in a conditional generative latent space. Overall, our method enables automatic construction of animatable human avatars for various types of clothes without the need for scanning subject-specific templates, and can generate realistic images with dynamic details for novel poses. Experiment show that our method outperforms state-of-the-art methods both qualitatively and quantitatively.

cs.CV↗

arXiv2022

DoubleField: Bridging the Neural Surface and Radiance Fields for High-fidelity Human Reconstruction and Rendering

We introduce DoubleField, a novel framework combining the merits of both surface field and radiance field for high-fidelity human reconstruction and rendering. Within DoubleField, the surface field and radiance field are associated together by a shared feature embedding and a surface-guided sampling strategy. Moreover, a view-to-view transformer is introduced to fuse multi-view features and learn view-dependent features directly from high-resolution inputs. With the modeling power of DoubleField and the view-to-view transformer, our method significantly improves the reconstruction quality of both geometry and appearance, while supporting direct inference, scene-specific high-resolution finetuning, and fast rendering. The efficacy of DoubleField is validated by the quantitative evaluations on several datasets and the qualitative results in a real-world sparse multi-view system, showing its superior capability for high-quality human model reconstruction and photo-realistic free-viewpoint human rendering. Data and source code will be made public for the research purpose. Please refer to our project page: http://www.liuyebin.com/dbfield/dbfield.html.

cs.CV↗

arXiv2022

Interacting Attention Graph for Single Image Two-Hand Reconstruction

Graph convolutional network (GCN) has achieved great success in single hand reconstruction task, while interacting two-hand reconstruction by GCN remains unexplored. In this paper, we present Interacting Attention Graph Hand (IntagHand), the first graph convolution based network that reconstructs two interacting hands from a single RGB image. To solve occlusion and interaction challenges of two-hand reconstruction, we introduce two novel attention based modules in each upsampling step of the original GCN. The first module is the pyramid image feature attention (PIFA) module, which utilizes multiresolution features to implicitly obtain vertex-to-image alignment. The second module is the cross hand attention (CHA) module that encodes the coherence of interacting hands by building dense cross-attention between two hand vertices. As a result, our model outperforms all existing two-hand reconstruction methods by a large margin on InterHand2.6M benchmark. Moreover, ablation studies verify the effectiveness of both PIFA and CHA modules for improving the reconstruction accuracy. Results on in-the-wild images and live video streams further demonstrate the generalization ability of our network. Our code is available at https://github.com/Dw1010/IntagHand.

cs.CV↗

arXiv2022

Salvaging Federated Learning by Local Adaptation

Federated learning (FL) is a heavily promoted approach for training ML models on sensitive data, e.g., text typed by users on their smartphones. FL is expressly designed for training on data that are unbalanced and non-iid across the participants. To ensure privacy and integrity of the fedeated model, latest FL approaches use differential privacy or robust aggregation. We look at FL from the \emph{local} viewpoint of an individual participant and ask: (1) do participants have an incentive to participate in FL? (2) how can participants \emph{individually} improve the quality of their local models, without re-designing the FL framework and/or involving other participants? First, we show that on standard tasks such as next-word prediction, many participants gain no benefit from FL because the federated model is less accurate on their data than the models they can train locally on their own. Second, we show that differential privacy and robust aggregation make this problem worse by further destroying the accuracy of the federated model for many participants. Then, we evaluate three techniques for local adaptation of federated models: fine-tuning, multi-task learning, and knowledge distillation. We analyze where each is applicable and demonstrate that all participants benefit from local adaptation. Participants whose local models are poor obtain big accuracy improvements over conventional FL. Participants whose local models are better than the federated model\textemdash and who have no incentive to participate in FL today\textemdash improve less, but sufficiently to make the adapted federated model better than their local models.

cs.LG↗

arXiv2022

Context-Based MEC Platform for Augmented-Reality Services in 5G Networks

Augmented reality (AR) has drawn great attention in recent years. However, current AR devices have drawbacks, e.g., weak computation ability and large power consumption. To solve the problem, mobile edge computing (MEC) can be introduced as a key technology to offload data and computation from AR devices to MEC servers via 5th Generation Mobile Communication Technology (5G) networks. To this end, a context-based MEC platform for AR services in 5G networks is proposed in this paper. On the platform, MEC is employed as a data processing center while AR devices are simplified as universal input/output devices, which overcomes their limitations and achieves better user experience. Moreover, the proof-of-concept (PoC) hardware prototype of the platform, and two typical use cases providing AR services of navigation and face recognition respectively are implemented to demonstrate the feasibility and effectiveness of the platform. Finally, the performance of the platform is also numerically evaluated, and the results validate the system design and agree well with the design expectations.

cs.NI↗

arXiv2022

Ground Experiment of Full-Duplex Multi-UAV System Enabled by Directional Antennas

A high performance multi-UAV communication system, which bridges multiple UAVs and ground station, is one of the key enablers to realize a variety of UAV-based systems. To address the issues such as the low spectrum efficiency caused by the co-channel interference, we have proposed a spectrum-efficient full-duplex multi-UA V communication system with low hardware complexity. In this paper, on-ground experiments are conducted to confirm the feasibility and effectiveness of the key feature of the proposed system, i.e., co-channel interference cancellation among UAVs by directional antennas and UAV position control, instead of energy-consuming dedicated self-interference cancellers on UAVs in traditional full-duplex systems. Channel power of interference link between a pair of two UAVs reusing the same channel is measured, and the achievable channel capacity is also measured by a prototype system implemented by software-defined radio devices. The results of different antennas and different antenna heights are also compared. The experimental results agree well with the designs and confirm the feasibility and effectiveness of the proposed system. This ground experiment is a work in progress to provide preliminary results for the multi-UAV-based experiments in the air in the future.

cs.IT↗

arXiv2022

Full-Duplex Aerial Communication System for Multiple UAVs with Directional Antennas

UAV-based wireless systems, such as wireless relay and remote sensing, have attracted great attentions from academia and industry. To realize them, a high-performance wireless aerial communication system, which bridges UAVs and ground stations, is one of the key enablers. However, there are still issues hindering its development, such as the severe co-channel interference among UAVs, and the limited payload/battery-life of UAVs. To address the challenges, we propose an aerial communication system which enables system-level full-duplex communication of multiple UAVs with lower hardware complexities than ideal full-duplex communication systems. In the proposed system, each channel is re-assigned to the uplink and downlink of a pair of UAVs, and each UAV employ a pair of separated channels for its uplink and downlink. The co-channel interference between UAVs that reuse same channels is eliminated by exploiting advantages of UAVs' maneuverability and high-gain directional antennas equipped in UAVs and ground stations, so that dedicated cancellers are not necessary in the proposed system. The system design and performance analysis are given, and the simulation results well agree with the designs.

cs.IT↗