arXiv subjects

Bowen Li

Publications and source records attributed to Bowen Li.

At least 199 records · Page 11Linked to original sources

arXiv2022

Dissipative soliton generation and real-time dynamics in microresonator-filtered fiber lasers

Optical frequency combs in microresonators (microcombs) have a wide range of applications in science and technology, due to its compact size and access to considerably larger comb spacing. Despite recent successes, the problems of self-starting, high mode efficiency as well as high output power have not been fully addressed for conventional soliton microcombs. Recent demonstration of laser cavity soliton microcombs by nesting a microresonator into a fiber cavity, shows great potential to solve the problems. Here we comprehensively study the dissipative soliton generation and interaction dynamics in a microresonator-filtered fiber laser in both theory and experiment. We first bring theoretical insight into the mode-locking principle, discuss the parameters effect on soliton properties and provide experimental guidelines for broadband soliton generation. We predict chirped bright dissipative soliton with flat-top spectral envelope in microresonators with normal dispersion, which is fundamentally infeasible for externally driven case. Furthermore, we experimentally achieve soliton microcombs with large bandwidth of ~10 nm and high mode efficiency of 90.7%. Finally, by taking advantage of an ultrahigh-speed time magnifier, we study the real-time soliton formation and interaction dynamics and experimentally observe soliton Newton's cradle. Our study will benefit the design of the novel, high-efficiency and self-starting microcombs for real-world applications.

physics.optics↗

arXiv2022

Memory-Driven Text-to-Image Generation

We introduce a memory-driven semi-parametric approach to text-to-image generation, which is based on both parametric and non-parametric techniques. The non-parametric component is a memory bank of image features constructed from a training set of images. The parametric component is a generative adversarial network. Given a new text description at inference time, the memory bank is used to selectively retrieve image features that are provided as basic information of target images, which enables the generator to produce realistic synthetic results. We also incorporate the content information into the discriminator, together with semantic features, allowing the discriminator to make a more reliable prediction. Experimental results demonstrate that the proposed memory-driven semi-parametric approach produces more realistic images than purely parametric approaches, in terms of both visual fidelity and text-image semantic consistency.

cs.CV↗

arXiv2022

Proton: Probing Schema Linking Information from Pre-trained Language Models for Text-to-SQL Parsing

The importance of building text-to-SQL parsers which can be applied to new databases has long been acknowledged, and a critical step to achieve this goal is schema linking, i.e., properly recognizing mentions of unseen columns or tables when generating SQLs. In this work, we propose a novel framework to elicit relational structures from large-scale pre-trained language models (PLMs) via a probing procedure based on Poincaré distance metric, and use the induced relations to augment current graph-based parsers for better schema linking. Compared with commonly-used rule-based methods for schema linking, we found that probing relations can robustly capture semantic correspondences, even when surface forms of mentions and entities differ. Moreover, our probing procedure is entirely unsupervised and requires no additional parameters. Extensive experiments show that our framework sets new state-of-the-art performance on three benchmarks. We empirically verify that our probing procedure can indeed find desired relational structures through qualitative analysis. Our code can be found at https://github.com/AlibabaResearch/DAMO-ConvAI.

cs.CL↗

arXiv2022

Siamese Object Tracking for Unmanned Aerial Vehicle: A Review and Comprehensive Analysis

Unmanned aerial vehicle (UAV)-based visual object tracking has enabled a wide range of applications and attracted increasing attention in the field of intelligent transportation systems because of its versatility and effectiveness. As an emerging force in the revolutionary trend of deep learning, Siamese networks shine in UAV-based object tracking with their promising balance of accuracy, robustness, and speed. Thanks to the development of embedded processors and the gradual optimization of deep neural networks, Siamese trackers receive extensive research and realize preliminary combinations with UAVs. However, due to the UAV's limited onboard computational resources and the complex real-world circumstances, aerial tracking with Siamese networks still faces severe obstacles in many aspects. To further explore the deployment of Siamese networks in UAV-based tracking, this work presents a comprehensive review of leading-edge Siamese trackers, along with an exhaustive UAV-specific analysis based on the evaluation using a typical UAV onboard processor. Then, the onboard tests are conducted to validate the feasibility and efficacy of representative Siamese trackers in real-world UAV deployment. Furthermore, to better promote the development of the tracking community, this work analyzes the limitations of existing Siamese trackers and conducts additional experiments represented by low-illumination evaluations. In the end, prospects for the development of Siamese tracking for UAV-based intelligent transportation systems are deeply discussed. The unified framework of leading-edge Siamese trackers, i.e., code library, and the results of their experimental evaluations are available at https://github.com/vision4robotics/SiameseTracking4UAV .

cs.CV↗

arXiv2022

Robotic Interestingness via Human-Informed Few-Shot Object Detection

Interestingness recognition is crucial for decision making in autonomous exploration for mobile robots. Previous methods proposed an unsupervised online learning approach that can adapt to environments and detect interesting scenes quickly, but lack the ability to adapt to human-informed interesting objects. To solve this problem, we introduce a human-interactive framework, AirInteraction, that can detect human-informed objects via few-shot online learning. To reduce the communication bandwidth, we first apply an online unsupervised learning algorithm on the unmanned vehicle for interestingness recognition and then only send the potential interesting scenes to a base-station for human inspection. The human operator is able to draw and provide bounding box annotations for particular interesting objects, which are sent back to the robot to detect similar objects via few-shot learning. Only using few human-labeled examples, the robot can learn novel interesting object categories during the mission and detect interesting scenes that contain the objects. We evaluate our method on various interesting scene recognition datasets. To the best of our knowledge, it is the first human-informed few-shot object detection framework for autonomous exploration.

cs.RO↗

arXiv2022

AirDet: Few-Shot Detection without Fine-tuning for Autonomous Exploration

Few-shot object detection has attracted increasing attention and rapidly progressed in recent years. However, the requirement of an exhaustive offline fine-tuning stage in existing methods is time-consuming and significantly hinders their usage in online applications such as autonomous exploration of low-power robots. We find that their major limitation is that the little but valuable information from a few support images is not fully exploited. To solve this problem, we propose a brand new architecture, AirDet, and surprisingly find that, by learning class-agnostic relation with the support images in all modules, including cross-scale object proposal network, shots aggregation module, and localization network, AirDet without fine-tuning achieves comparable or even better results than many fine-tuned methods, reaching up to 30-40% improvements. We also present solid results of onboard tests on real-world exploration data from the DARPA Subterranean Challenge, which strongly validate the feasibility of AirDet in robotics. To the best of our knowledge, AirDet is the first feasible few-shot detection method for autonomous exploration of low-power robots. The code and pre-trained models are released at https://github.com/Jaraxxus-Me/AirDet.

cs.CV↗

arXiv2022

Universal Learned Image Compression With Low Computational Cost

Recently, learned image compression methods have developed rapidly and exhibited excellent rate-distortion performance when compared to traditional standards, such as JPEG, JPEG2000 and BPG. However, the learning-based methods suffer from high computational costs, which is not beneficial for deployment on devices with limited resources. To this end, we propose shift-addition parallel modules (SAPMs), including SAPM-E for the encoder and SAPM-D for the decoder, to largely reduce the energy consumption. To be specific, they can be taken as plug-and-play components to upgrade existing CNN-based architectures, where the shift branch is used to extract large-grained features as compared to small-grained features learned by the addition branch. Furthermore, we thoroughly analyze the probability distribution of latent representations and propose to use Laplace Mixture Likelihoods for more accurate entropy estimation. Experimental results demonstrate that the proposed methods can achieve comparable or even better performance on both PSNR and MS-SSIM metrics to that of the convolutional counterpart with an about 2x energy reduction.

eess.IV↗

arXiv2022

Ultrahigh ion diffusion in oxide crystal by engineering the interfacial transporter channels

The mass storage and removal in solid conductors always played vital role on the technological applications such as modern batteries, permeation membranes and neuronal computations, which were seriously lying on the ion diffusion and kinetics in bulk lattice. However, the ions transport was kinetically limited by the low diffusional process, which made it a challenge to fabricate applicable conductors with high electronic and ionic conductivities at room temperature. It was known that at essentially all interfaces, the existed space charge layers could modify the charge transport, storage and transfer properties. Thus, in the current study, we proposed an acid solution/WO3/ITO structure and achieved an ultrafast hydrogen transport in WO3 layer by interfacial job-sharing diffusion. In this sandwich structure, the transport pathways of the protons and electrons were spatially separated in acid solution and ITO layer respectively, resulting the pronounced increasing of effective hydrogen diffusion coefficient (Deff) up to 106 times. The experiment and theory simulations also revealed that this accelerated hydrogen transport based on the interfacial job-sharing diffusion was universal and could be extended to other ions and oxide materials as well, which would potentially stimulate systematic studies on ultrafast mixed conductors or faster solid-state electrochemical switching devices in the future.

cond-mat.mtrl-sci↗

arXiv2022

Unveiling the relative timing jitter in counter propagating all normal dispersion (CANDi) dual-comb fiber laser

Counter-propagating all-normal dispersion (CANDi) fiber laser is an emerging high-energy single-cavity dual-comb laser source. Its relative timing jitter (RTJ), a critical parameter for dual-comb timing precision and spectral resolution, has not been comprehensively investigated. In this paper, we enhance the state-of-the-art CANDi fiber laser pulse energy from 1 nJ to 8 nJ. We then introduce a novel reference-free RTJ characterization technique that provides shot-to-shot measurement capability at femtosecond precision for the first time. The measurement noise floor reaches 1.6x10-7 fs2/Hz, and the corresponding integrated measurement precision is only 1.8 fs (1 kHz, 20 MHz). With this new characterization tool, we are able to study the physical origin of CANDi laser's RTJ in detail. We first verify that the cavity length fluctuation does not contribute to the RTJ. Then we measure the integrated RTJ to be 39 fs (1 kHz, 20 MHz) and identify the pump relative intensity noise (RIN) to be the dominant factor responsible for it. In particular, pump RIN is coupled to the RTJ through the Gordon-Haus effect. Finally, solutions to reduce the free-running CANDi laser's RTJ are discussed. This work provides a general guideline to improve the performance of compact single-cavity dual-comb systems like CANDi laser benefitting various dual-comb applications.

physics.optics↗

arXiv2022

Correlation Filters for Unmanned Aerial Vehicle-Based Aerial Tracking: A Review and Experimental Evaluation

Aerial tracking, which has exhibited its omnipresent dedication and splendid performance, is one of the most active applications in the remote sensing field. Especially, unmanned aerial vehicle (UAV)-based remote sensing system, equipped with a visual tracking approach, has been widely used in aviation, navigation, agriculture,transportation, and public security, etc. As is mentioned above, the UAV-based aerial tracking platform has been gradually developed from research to practical application stage, reaching one of the main aerial remote sensing technologies in the future. However, due to the real-world onerous situations, e.g., harsh external challenges, the vibration of the UAV mechanical structure (especially under strong wind conditions), the maneuvering flight in complex environment, and the limited computation resources onboard, accuracy, robustness, and high efficiency are all crucial for the onboard tracking methods. Recently, the discriminative correlation filter (DCF)-based trackers have stood out for their high computational efficiency and appealing robustness on a single CPU, and have flourished in the UAV visual tracking community. In this work, the basic framework of the DCF-based trackers is firstly generalized, based on which, 23 state-of-the-art DCF-based trackers are orderly summarized according to their innovations for solving various issues. Besides, exhaustive and quantitative experiments have been extended on various prevailing UAV tracking benchmarks, i.e., UAV123, UAV123@10fps, UAV20L, UAVDT, DTB70, and VisDrone2019-SOT, which contain 371,903 frames in total. The experiments show the performance, verify the feasibility, and demonstrate the current challenges of DCF-based trackers onboard UAV tracking.

cs.CV↗

arXiv2022

A Smart Contract based Crowdfunding Mechanism for Hierarchical Federated Learning

Hierarchical Federated Learning (HFL) is introduced as a promising technique that allows model owners to fully exploit computational resources and bandwidth resources to train the global model. However, due to the high training cost, a single model owner may not be able to deploy HFL. To address this issue, we develop a smart contract based trust crowdfunding mechanism for HFL, which enables multiple model owners to obtain a crowdfunding model with high social utility for multiple crowdfunding participants. To ensure the authenticity of the crowdfunding mechanism, we implemented the Vickey-Clark-Croves (VCG) mechanism to encourage all crowdfunding participants and clients to provide realistic bids and offers. At the same time, in order to ensure guaranteed trustworthiness of crowdfunding and automatic distribution of funds, we develop and implement a smart contract to record the crowdfunding process and training results in the blockchain. We prove that the proposed scheme satisfies the budget balance and participant constraint. Finally, we implement a prototype of this smart contract on an Ethereoum private chain and evaluate the proposed VCG mechanism. The experimental results demonstrate that the proposed scheme can effectively improve social utility while ensuring the authenticity and trustworthiness of the crowdfunding process.

cs.DC↗

arXiv2022

Experimental Realization of the Rabi-Hubbard Model with Trapped Ions

Quantum simulation provides important tools in studying strongly correlated many-body systems with controllable parameters. As a hybrid of two fundamental models in quantum optics and in condensed matter physics, the Rabi-Hubbard model demonstrates rich physics through the competition between local spin-boson interactions and long-range boson hopping. Here we report an experimental realization of the Rabi-Hubbard model using up to $16$ trapped ions and present a controlled study of its equilibrium properties and quantum dynamics. We observe the ground-state quantum phase transition by slowly quenching the coupling strength, and measure the quantum dynamical evolution in various parameter regimes. With the magnetization and the spin-spin correlation as probes, we verify the prediction of the model Hamiltonian by comparing theoretical results in small system sizes with experimental observations. For larger-size systems of $16$ ions and $16$ phonon modes, the effective Hilbert space dimension exceeds $2^{57}$, whose dynamics is intractable for classical supercomputers.

quant-ph↗

arXiv2022

Blindly Assess Quality of In-the-Wild Videos via Quality-aware Pre-training and Motion Perception

Perceptual quality assessment of the videos acquired in the wilds is of vital importance for quality assurance of video services. The inaccessibility of reference videos with pristine quality and the complexity of authentic distortions pose great challenges for this kind of blind video quality assessment (BVQA) task. Although model-based transfer learning is an effective and efficient paradigm for the BVQA task, it remains to be a challenge to explore what and how to bridge the domain shifts for better video representation. In this work, we propose to transfer knowledge from image quality assessment (IQA) databases with authentic distortions and large-scale action recognition with rich motion patterns. We rely on both groups of data to learn the feature extractor. We train the proposed model on the target VQA databases using a mixed list-wise ranking loss function. Extensive experiments on six databases demonstrate that our method performs very competitively under both individual database and mixed database training settings. We also verify the rationality of each component of the proposed method and explore a simple manner for further improvement.

eess.IV↗

arXiv2022

DarkLighter: Light Up the Darkness for UAV Tracking

Recent years have witnessed the fast evolution and promising performance of the convolutional neural network (CNN)-based trackers, which aim at imitating biological visual systems. However, current CNN-based trackers can hardly generalize well to low-light scenes that are commonly lacked in the existing training set. In indistinguishable night scenarios frequently encountered in unmanned aerial vehicle (UAV) tracking-based applications, the robustness of the state-of-the-art (SOTA) trackers drops significantly. To facilitate aerial tracking in the dark through a general fashion, this work proposes a low-light image enhancer namely DarkLighter, which dedicates to alleviate the impact of poor illumination and noise iteratively. A lightweight map estimation network, i.e., ME-Net, is trained to efficiently estimate illumination maps and noise maps jointly. Experiments are conducted with several SOTA trackers on numerous UAV dark tracking scenes. Exhaustive evaluations demonstrate the reliability and universality of DarkLighter, with high efficiency. Moreover, DarkLighter has further been implemented on a typical UAV system. Real-world tests at night scenes have verified its practicability and dependability.

cs.CV↗

arXiv2022

S$^2$SQL: Injecting Syntax to Question-Schema Interaction Graph Encoder for Text-to-SQL Parsers

The task of converting a natural language question into an executable SQL query, known as text-to-SQL, is an important branch of semantic parsing. The state-of-the-art graph-based encoder has been successfully used in this task but does not model the question syntax well. In this paper, we propose S$^2$SQL, injecting Syntax to question-Schema graph encoder for Text-to-SQL parsers, which effectively leverages the syntactic dependency information of questions in text-to-SQL to improve the performance. We also employ the decoupling constraint to induce diverse relational edge embedding, which further improves the network's performance. Experiments on the Spider and robustness setting Spider-Syn demonstrate that the proposed approach outperforms all existing methods when pre-training models are used, resulting in a performance ranks first on the Spider leaderboard.

cs.CL↗

arXiv2021

Ferroelectric-HfO2/Oxide Interfaces, Oxygen Distribution Effect and Implications for Device Performance

Atomic-scale understanding of HfO2 ferroelectricity is important to help address many challenges in developing reliable and high-performance ferroelectric HfO2 (fe-HfO2) based devices. Though investigated from different angles, a factor that is real device-relevant and clearly deserves more attention has largely been overlooked by previous research, namely, the fe-HfO2/dielectric interface. Here, we investigate the electronic structures of several typical interfaces formed between ultrathin fe-HfO2 and oxide dielectrics in the sub-3-nm region. We find that interface formation introduces strong depolarizing fields in fe-HfO2, which is detrimental for ferroelectric polarization but can be a merit if tamed for tunneling devices, as recently demonstrated. Asymmetric oxygen distribution-induced polarity, intertwined with ferroelectric polarization or not, is also investigated as a relevant interfacial effect in real device. Though considered detrimental from certain aspects, such as inducing build-in field (independent of ferroelectric polarization) and exacerbating depolarization (intertwined with ferroelectric polarization), it can be partly balanced out by other effects, such as annealing (extrinsic) and polarity-induced defect formation (intrinsic). This work provides insights into ferroelectric-HfO2/dielectric interfaces and some useful implications for the development of devices.

cond-mat.mtrl-sci↗

arXiv2021

Deep-3D Microscope: 3D volumetric microscopy of thick scattering samples using a wide-field microscope and machine learning

Confocal microscopy is the standard approach for obtaining volumetric images of a sample with high axial and lateral resolution, especially when dealing with scattering samples. Unfortunately, a confocal microscope is quite expensive compared to traditional microscopes. In addition, the point scanning in a confocal leads to slow imaging speed and photobleaching due to the high dose of laser energy. In this paper, we demonstrate how the advances in machine learning can be exploited to "teach" a traditional wide-field microscope, one that's available in every lab, into producing 3D volumetric images like a confocal. The key idea is to obtain multiple images with different focus settings using a wide-field microscope and use a 3D Generative Adversarial Network (GAN) based neural network to learn the mapping between the blurry low-contrast image stack obtained using wide-field and the sharp, high-contrast images obtained using a confocal. After training the network with widefield-confocal image pairs, the network can reliably and accurately reconstruct 3D volumetric images that rival confocal in terms of its lateral resolution, z-sectioning and image contrast. Our experimental results demonstrate generalization ability to handle unseen data, stability in the reconstruction results, high spatial resolution even when imaging thick ($\sim40$ microns) highly-scattering samples. We believe that such learning-based-microscopes have the potential to bring confocal quality imaging to every lab that has a wide-field microscope.

physics.optics↗

arXiv2021

Learning from Subjective Ratings Using Auto-Decoded Deep Latent Embeddings

Depending on the application, radiological diagnoses can be associated with high inter- and intra-rater variabilities. Most computer-aided diagnosis (CAD) solutions treat such data as incontrovertible, exposing learning algorithms to considerable and possibly contradictory label noise and biases. Thus, managing subjectivity in labels is a fundamental problem in medical imaging analysis. To address this challenge, we introduce auto-decoded deep latent embeddings (ADDLE), which explicitly models the tendencies of each rater using an auto-decoder framework. After a simple linear transformation, the latent variables can be injected into any backbone at any and multiple points, allowing the model to account for rater-specific effects on the diagnosis. Importantly, ADDLE does not expect multiple raters per image in training, meaning it can readily learn from data mined from hospital archives. Moreover, the complexity of training ADDLE does not increase as more raters are added. During inference each rater can be simulated and a 'mean' or 'greedy' virtual rating can be produced. We test ADDLE on the problem of liver steatosis diagnosis from 2D ultrasound (US) by collecting 46 084 studies along with clinical US diagnoses originating from 65 different raters. We evaluated diagnostic performance using a separate dataset with gold-standard biopsy diagnoses. ADDLE can improve the partial areas under the curve (AUCs) for diagnosing severe steatosis by 10.5% over standard classifiers while outperforming other annotator-noise approaches, including those requiring 65 times the parameters.

cs.CV↗