arXiv subjects

Xinyu Zhang

Publications and source records attributed to Xinyu Zhang.

At least 379 records · Page 21Linked to original sources

arXiv2022

Coupled Modeling and Fusion Control for a Multi-modal Deformable Land-air Robot

A deformable land-air robot is introduced with excellent driving and flying capabilities, offering a smooth switching mechanism between the two modes. An elaborate coupled dynamics model is established for the robot, including rotors, chassis, suspension, and the deformable structure. In addition, a model-based controller is designed for landing and mode switching in various unstructured conditions, such as slopes and curved surface. And considering locomotion and complex near-ground situations to achieve cooperation between the two fused modalities. This system was simulated in ADAMS/Simulink and a tested with hardware-in-the-loop system was constructed for testing in various slopes. With a designed controller, the results showed the robot is capable of fast and smooth land-air switching, with a 24.6 % faster landing on slopes. The controller can also reduce landing offset and impact force more effectively than the normal control method at 32.7 % and 34.3 %, respectively.

cs.RO↗

arXiv2022

Asymptotic Properties of the Synthetic Control Method

This paper provides new insights into the asymptotic properties of the synthetic control method (SCM). We show that the synthetic control (SC) weight converges to a limiting weight that minimizes the mean squared prediction risk of the treatment-effect estimator when the number of pretreatment periods goes to infinity, and we also quantify the rate of convergence. Observing the link between the SCM and model averaging, we further establish the asymptotic optimality of the SC estimator under imperfect pretreatment fit, in the sense that it achieves the lowest possible squared prediction error among all possible treatment effect estimators that are based on an average of control units, such as matching, inverse probability weighting and difference-in-differences. The asymptotic optimality holds regardless of whether the number of control units is fixed or divergent. Thus, our results provide justifications for the SCM in a wide range of applications. The theoretical results are verified via simulations.

econ.EM↗

arXiv2022

CAE v2: Context Autoencoder with CLIP Target

Masked image modeling (MIM) learns visual representation by masking and reconstructing image patches. Applying the reconstruction supervision on the CLIP representation has been proven effective for MIM. However, it is still under-explored how CLIP supervision in MIM influences performance. To investigate strategies for refining the CLIP-targeted MIM, we study two critical elements in MIM, i.e., the supervision position and the mask ratio, and reveal two interesting perspectives, relying on our developed simple pipeline, context autodecoder with CLIP target (CAE v2). Firstly, we observe that the supervision on visible patches achieves remarkable performance, even better than that on masked patches, where the latter is the standard format in the existing MIM methods. Secondly, the optimal mask ratio positively correlates to the model size. That is to say, the smaller the model, the lower the mask ratio needs to be. Driven by these two discoveries, our simple and concise approach CAE v2 achieves superior performance on a series of downstream tasks. For example, a vanilla ViT-Large model achieves 81.7% and 86.7% top-1 accuracy on linear probing and fine-tuning on ImageNet-1K, and 55.9% mIoU on semantic segmentation on ADE20K with the pre-training for 300 epochs. We hope our findings can be helpful guidelines for the pre-training in the MIM area, especially for the small-scale models.

cs.CV↗

arXiv2022

A classification performance evaluation measure considering data separability

Machine learning and deep learning classification models are data-driven, and the model and the data jointly determine their classification performance. It is biased to evaluate the model's performance only based on the classifier accuracy while ignoring the data separability. Sometimes, the model exhibits excellent accuracy, which might be attributed to its testing on highly separable data. Most of the current studies on data separability measures are defined based on the distance between sample points, but this has been demonstrated to fail in several circumstances. In this paper, we propose a new separability measure--the rate of separability (RS), which is based on the data coding rate. We validate its effectiveness as a supplement to the separability measure by comparing it to four other distance-based measures on synthetic datasets. Then, we demonstrate the positive correlation between the proposed measure and recognition accuracy in a multi-task scenario constructed from a real dataset. Finally, we discuss the methods for evaluating the classification performance of machine learning and deep learning models considering data separability.

cs.LG↗

arXiv2022

3DFill:Reference-guided Image Inpainting by Self-supervised 3D Image Alignment

Most existing image inpainting algorithms are based on a single view, struggling with large holes or the holes containing complicated scenes. Some reference-guided algorithms fill the hole by referring to another viewpoint image and use 2D image alignment. Due to the camera imaging process, simple 2D transformation is difficult to achieve a satisfactory result. In this paper, we propose 3DFill, a simple and efficient method for reference-guided image inpainting. Given a target image with arbitrary hole regions and a reference image from another viewpoint, the 3DFill first aligns the two images by a two-stage method: 3D projection + 2D transformation, which has better results than 2D image alignment. The 3D projection is an overall alignment between images and the 2D transformation is a local alignment focused on the hole region. The entire process of image alignment is self-supervised. We then fill the hole in the target image with the contents of the aligned image. Finally, we use a conditional generation network to refine the filled image to obtain the inpainting result. 3DFill achieves state-of-the-art performance on image inpainting across a variety of wide view shifts and has a faster inference speed than other inpainting models.

cs.CV↗

arXiv2022

Measuring charge distribution of molecular cations by atomic Coulomb probe microscope

Imaging the charge distributions and structures of molecules and clusters will promote the understanding of the dynamics of the quantum system. Here, we report a method by using an Ar atom as a tip to probe the charge distributions of benzene (Bz) cations in gas phase. Remarkably, the measured charge distributions of Bz cation (QH =0.204,QC=-0.037)and dication (QH =0.248,QC=0.0853)agree well with the calculated Mulliken distributions,and the structures of Bz dimer is reconstructed by using the measured charge distributions. The structures of two Bz dimer isomers (T-shaped and PD isomers) can be resolved from the measured inter-molecular potential V(R) between two Bz ions, and the structures of Bz dimer agree well with the theoretical predictions.

physics.atm-clus↗

arXiv2022

RF-CHORD: Towards Deployable RFID Localization System for Logistics Network

RFID localization is considered the key enabler of automating the process of inventory tracking and management for high-performance logistic network. A practical and deployable RFID localization system needs to meet reliability, throughput, and range requirements. This paper presents RF-Chord, the first RFID localization system that simultaneously meets all three requirements. RF-Chord features a one-shot multisine-constructed wideband design that can process RF signal with a 200 MHz bandwidth in real-time to facilitate one-shot localization at scale. In addition, multiple SINR enhancement techniques are designed for range extension. On top of that, a kernel-layer-based near-field localization framework and a multipath-suppression algorithm are proposed to reduce the 99% long-tail errors. Our empirical results show that RF-Chord can localize more than 180 tags 6 m away from a reader within 1 second and with 99% long-tail error of 0.786 m, achieving a 0% miss reading rate and ~0.01% cross-reading rate in the warehouse and fresh food delivery store deployment.

eess.SP↗

arXiv2022

MTCSNN: Multi-task Clinical Siamese Neural Network for Diabetic Retinopathy Severity Prediction

Diabetic Retinopathy (DR) has become one of the leading causes of vision impairment in working-aged people and is a severe problem worldwide. However, most of the works ignored the ordinal information of labels. In this project, we propose a novel design MTCSNN, a Multi-task Clinical Siamese Neural Network for Diabetic Retinopathy severity prediction task. The novelty of this project is to utilize the ordinal information among labels and add a new regression task, which can help the model learn more discriminative feature embedding for fine-grained classification tasks. We perform comprehensive experiments over the RetinaMNIST, comparing MTCSNN with other models like ResNet-18, 34, 50. Our results indicate that MTCSNN outperforms the benchmark models in terms of AUC and accuracy on the test dataset.

cs.CV↗

arXiv2022

Making a MIRACL: Multilingual Information Retrieval Across a Continuum of Languages

MIRACL (Multilingual Information Retrieval Across a Continuum of Languages) is a multilingual dataset we have built for the WSDM 2023 Cup challenge that focuses on ad hoc retrieval across 18 different languages, which collectively encompass over three billion native speakers around the world. These languages have diverse typologies, originate from many different language families, and are associated with varying amounts of available resources -- including what researchers typically characterize as high-resource as well as low-resource languages. Our dataset is designed to support the creation and evaluation of models for monolingual retrieval, where the queries and the corpora are in the same language. In total, we have gathered over 700k high-quality relevance judgments for around 77k queries over Wikipedia in these 18 languages, where all assessments have been performed by native speakers hired by our team. Our goal is to spur research that will improve retrieval across a continuum of languages, thus enhancing information access capabilities for diverse populations around the world, particularly those that have been traditionally underserved. This overview paper describes the dataset and baselines that we share with the community. The MIRACL website is live at http://miracl.ai/.

cs.IR↗

arXiv2022

Better Than Whitespace: Information Retrieval for Languages without Custom Tokenizers

Tokenization is a crucial step in information retrieval, especially for lexical matching algorithms, where the quality of indexable tokens directly impacts the effectiveness of a retrieval system. Since different languages have unique properties, the design of the tokenization algorithm is usually language-specific and requires at least some lingustic knowledge. However, only a handful of the 7000+ languages on the planet benefit from specialized, custom-built tokenization algorithms, while the other languages are stuck with a "default" whitespace tokenizer, which cannot capture the intricacies of different languages. To address this challenge, we propose a different approach to tokenization for lexical matching retrieval algorithms (e.g., BM25): using the WordPiece tokenizer, which can be built automatically from unsupervised data. We test the approach on 11 typologically diverse languages in the MrTyDi collection: results show that the mBERT tokenizer provides strong relevance signals for retrieval "out of the box", outperforming whitespace tokenization on most languages. In many cases, our approach also improves retrieval effectiveness when combined with existing custom-built tokenizers.

cs.CL↗

arXiv2022

BYHE: A Simple Framework for Boosting End-to-end Video-based Heart Rate Measurement Network

Heart rate measuring based on remote photoplethysmography (rPPG) plays an important role in health caring, which estimates heart rate from facial video in a non-contact, less-constrained way. End-to-end neural network is a main branch of rPPG-based heart rate estimation methods, whose trait is recovering rPPG signal containing sufficient heart rate message from original facial video directly. However, there exists some easily neglected problems on relevant datasets which thwarting the efficient training of end-to-end methods, such as uncertain temporal delay and indefinite envelope shape of label waves. Although many novel and powerful networks are proposed, hitherto there are no systematic research digging into these problems. In this paper, from perspective of common intrinsic rhythm periodical self-similarity results from cardiac activities, we propose a comprehensive methodology, Boost Your Heartbeat Estimation (BYHE), including new label representations, corresponding network adjustments and loss functions. BYHE can be easily grafted on current end-to-end network and boost its training efficiency. By applying our methodology, we can save tremendous time without conducting laborious handworks, such as label wave alignment which is necessary for previous end-to-end methods, and meanwhile enhance the utilization on datasets. According to our experiments, BYHE can leverage classical end-to-end network to reach competitive performance against those state-of-the-art methods on mostly used datasets. Such improvement indicates selecting perspicuous and efficient label representation is also a promising direction towards better remote physiological signal measurement.

cs.CV↗

arXiv2022

From Easy to Hard: A Dual Curriculum Learning Framework for Context-Aware Document Ranking

Contextual information in search sessions is important for capturing users' search intents. Various approaches have been proposed to model user behavior sequences to improve document ranking in a session. Typically, training samples of (search context, document) pairs are sampled randomly in each training epoch. In reality, the difficulty to understand user's search intent and to judge document's relevance varies greatly from one search context to another. Mixing up training samples of different difficulties may confuse the model's optimization process. In this work, we propose a curriculum learning framework for context-aware document ranking, in which the ranking model learns matching signals between the search context and the candidate document in an easy-to-hard manner. In so doing, we aim to guide the model gradually toward a global optimum. To leverage both positive and negative examples, two curricula are designed. Experiments on two real query log datasets show that our proposed framework can improve the performance of several existing methods significantly, demonstrating the effectiveness of curriculum learning for context-aware document ranking.

cs.IR↗

arXiv2022

Pre-training for Information Retrieval: Are Hyperlinks Fully Explored?

Recent years have witnessed great progress on applying pre-trained language models, e.g., BERT, to information retrieval (IR) tasks. Hyperlinks, which are commonly used in Web pages, have been leveraged for designing pre-training objectives. For example, anchor texts of the hyperlinks have been used for simulating queries, thus constructing tremendous query-document pairs for pre-training. However, as a bridge across two web pages, the potential of hyperlinks has not been fully explored. In this work, we focus on modeling the relationship between two documents that are connected by hyperlinks and designing a new pre-training objective for ad-hoc retrieval. Specifically, we categorize the relationships between documents into four groups: no link, unidirectional link, symmetric link, and the most relevant symmetric link. By comparing two documents sampled from adjacent groups, the model can gradually improve its capability of capturing matching signals. We propose a progressive hyperlink predication ({PHP}) framework to explore the utilization of hyperlinks in pre-training. Experimental results on two large-scale ad-hoc retrieval datasets and six question-answering datasets demonstrate its superiority over existing pre-training methods.

cs.IR↗

arXiv2022

CAMO-MOT: Combined Appearance-Motion Optimization for 3D Multi-Object Tracking with Camera-LiDAR Fusion

3D Multi-object tracking (MOT) ensures consistency during continuous dynamic detection, conducive to subsequent motion planning and navigation tasks in autonomous driving. However, camera-based methods suffer in the case of occlusions and it can be challenging to accurately track the irregular motion of objects for LiDAR-based methods. Some fusion methods work well but do not consider the untrustworthy issue of appearance features under occlusion. At the same time, the false detection problem also significantly affects tracking. As such, we propose a novel camera-LiDAR fusion 3D MOT framework based on the Combined Appearance-Motion Optimization (CAMO-MOT), which uses both camera and LiDAR data and significantly reduces tracking failures caused by occlusion and false detection. For occlusion problems, we are the first to propose an occlusion head to select the best object appearance features multiple times effectively, reducing the influence of occlusions. To decrease the impact of false detection in tracking, we design a motion cost matrix based on confidence scores which improve the positioning and object prediction accuracy in 3D space. As existing multi-object tracking methods only consider a single category, we also propose to build a multi-category loss to implement multi-object tracking in multi-category scenes. A series of validation experiments are conducted on the KITTI and nuScenes tracking benchmarks. Our proposed method achieves state-of-the-art performance and the lowest identity switches (IDS) value (23 for Car and 137 for Pedestrian) among all multi-modal MOT methods on the KITTI test dataset. And our proposed method achieves state-of-the-art performance among all algorithms on the nuScenes test dataset with 75.3% AMOTA.

cs.CV↗

arXiv2022

Ambiguity Function Shaping based on Alternating Direction Riemannian Optimal Algorithm

In order to improve the ability of cognitive radar (CR) to adapt to the environment, the required ambiguity function (AF) can be synthesized by designing the waveform. The key to this problem is how to minimize the interference power. Suppressing the interference power is equivalent to minimize the expectation of slow-time ambiguity function (STAF) over range-Doppler bins. From a technical point of view, this is actually an optimization problem of a non-convex quartic function with constant modulus constraints (CMC). In this paper, we proposed a novel method to design a waveform to synthesize the STAF based on suppressing the interference power. We put forward an iterative algorithm within an alternating direction penalty method (ADPM) framework. In each iteration, this problem is split into two sub-problems by introducing auxiliary variables. In the first sub-problem, we solved the convex problem directly with a closed-form solution, then utilized the Riemannian trust region (RTR) algorithm in the second sub-problem. Simulation results demonstrate that the proposed algorithm outperforms other advanced algorithms in the aspects of STAF, range-cut and signal-to-interference-ratio (SIR) value.

eess.SP↗

arXiv2022

Coarse-to-Fine: Hierarchical Multi-task Learning for Natural Language Understanding

Generalized text representations are the foundation of many natural language understanding tasks. To fully utilize the different corpus, it is inevitable that models need to understand the relevance among them. However, many methods ignore the relevance and adopt a single-channel model (a coarse paradigm) directly for all tasks, which lacks enough rationality and interpretation. In addition, some existing works learn downstream tasks by stitches skill block(a fine paradigm), which might cause irrationalresults due to its redundancy and noise. Inthis work, we first analyze the task correlation through three different perspectives, i.e., data property, manual design, and model-based relevance, based on which the similar tasks are grouped together. Then, we propose a hierarchical framework with a coarse-to-fine paradigm, with the bottom level shared to all the tasks, the mid-level divided to different groups, and the top-level assigned to each of the tasks. This allows our model to learn basic language properties from all tasks, boost performance on relevant tasks, and reduce the negative impact from irrelevant tasks. Our experiments on 13 benchmark datasets across five natural language understanding tasks demonstrate the superiority of our method.

cs.CL↗

arXiv2022

Functionally Reconfigurable Integrated Structure of Fabry-Perot Antenna and Wideband Liquid Absorber

This paper proposes a functionally reconfigurable integrated structure of a Fabry-Perot (FP) antenna and wideband liquid absorber. The partial reflecting surface (PRS) is elaborately tailored to serve as both a component of the FP antenna and the metal ground of the broadband liquid absorber. Then the integrated structure is realized by combining the FP antenna with the liquid absorber. The PRS is composed of patches on the top layer of the substrate and the square loop on the bottom and the microstrip patch antenna is used as a source. The liquid absorber is composed of a 3-D printed container, 45% ethanol layer, and a metal ground. By this design, the structure can serve as the FP antenna when the ethanol is extracted or the liquid absorber when the ethanol is injected. This can be used to switch between detection and stealth mode which can better adapt to the complex electromagnetic environment. The simulation results show that the gain of the antenna reaches to 19.7 dBi and the operating band is 100 MHz when the ethanol is extracted. When the ethanol is injected, the absorption band of the absorber ranges from 4 GHz to 20 GHz and the absorption bandwidth is over 133%. The monostatic RCS reduction bands of the structure with ethanol range from 4 GHz to 20GHz and the average RCS reduction is 28.4 dB. The measured and simulated results agree well.

physics.app-ph↗

arXiv2022

Intelligent Amphibious Ground-Aerial Vehicles: State of the Art Technology for Future Transportation

Amphibious ground-aerial vehicles fuse flying and driving modes to enable more flexible air-land mobility and have received growing attention recently. By analyzing the existing amphibious vehicles, we highlight the autonomous fly-driving functionality for the effective uses of amphibious vehicles in complex three-dimensional urban transportation systems. We review and summarize the key enabling technologies for intelligent flying-driving in existing amphibious vehicle designs, identify major technological barriers and propose potential solutions for future research and innovation. This paper aims to serve as a guide for research and development of intelligent amphibious vehicles for urban transportation toward the future.

cs.RO↗