arXiv subjects

Guofeng Zhang

Publications and source records attributed to Guofeng Zhang.

At least 181 records · Page 10Linked to original sources

arXiv2022

Robust Self-Supervised LiDAR Odometry via Representative Structure Discovery and 3D Inherent Error Modeling

The correct ego-motion estimation basically relies on the understanding of correspondences between adjacent LiDAR scans. However, given the complex scenarios and the low-resolution LiDAR, finding reliable structures for identifying correspondences can be challenging. In this paper, we delve into structure reliability for accurate self-supervised ego-motion estimation and aim to alleviate the influence of unreliable structures in training, inference and mapping phases. We improve the self-supervised LiDAR odometry substantially from three aspects: 1) A two-stage odometry estimation network is developed, where we obtain the ego-motion by estimating a set of sub-region transformations and averaging them with a motion voting mechanism, to encourage the network focusing on representative structures. 2) The inherent alignment errors, which cannot be eliminated via ego-motion optimization, are down-weighted in losses based on the 3D point covariance estimations. 3) The discovered representative structures and learned point covariances are incorporated in the mapping module to improve the robustness of map construction. Our two-frame odometry outperforms the previous state of the arts by 16%/12% in terms of translational/rotational errors on the KITTI dataset and performs consistently well on the Apollo-Southbay datasets. We can even rival the fully supervised counterparts with our mapping module and more unlabeled training data.

cs.CV↗

arXiv2022

SelfVoxeLO: Self-supervised LiDAR Odometry with Voxel-based Deep Neural Networks

Recent learning-based LiDAR odometry methods have demonstrated their competitiveness. However, most methods still face two substantial challenges: 1) the 2D projection representation of LiDAR data cannot effectively encode 3D structures from the point clouds; 2) the needs for a large amount of labeled data for training limit the application scope of these methods. In this paper, we propose a self-supervised LiDAR odometry method, dubbed SelfVoxeLO, to tackle these two difficulties. Specifically, we propose a 3D convolution network to process the raw LiDAR data directly, which extracts features that better encode the 3D geometric patterns. To suit our network to self-supervised learning, we design several novel loss functions that utilize the inherent properties of LiDAR point clouds. Moreover, an uncertainty-aware mechanism is incorporated in the loss functions to alleviate the interference of moving objects/noises. We evaluate our method's performances on two large-scale datasets, i.e., KITTI and Apollo-SouthBay. Our method outperforms state-of-the-art unsupervised methods by 27%/32% in terms of translational/rotational errors on the KITTI dataset and also performs well on the Apollo-SouthBay dataset. By including more unlabelled training data, our method can further improve performance comparable to the supervised methods.

cs.CV↗

arXiv2022

Efficient Quantum Feature Extraction for CNN-based Learning

Recent work has begun to explore the potential of parametrized quantum circuits (PQCs) as general function approximators. In this work, we propose a quantum-classical deep network structure to enhance classical CNN model discriminability. The convolutional layer uses linear filters to scan the input data. Moreover, we build PQC, which is a more potent function approximator, with more complex structures to capture the features within the receptive field. The feature maps are obtained by sliding the PQCs over the input in a similar way as CNN. We also give a training algorithm for the proposed model. The hybrid models used in our design are validated by numerical simulation. We demonstrate the reasonable classification performances on MNIST and we compare the performances with models in different settings. The results disclose that the model with ansatz in high expressibility achieves lower cost and higher accuracy.

quant-ph↗

arXiv2022

Amplification of optical Schrödinger cat states with implementation protocol based on frequency comb

We proposed and analyzed a scheme to generate large-size Schrödinger cat states based on linear operations of Fock states and squeezed vacuum states and conditional measurements. By conducting conditional measurements via photon number detectors, two unbalanced Schrödinger kitten states combined by a beam splitter can be amplified to a large-size cat state with the same parity. According to simulation results, two Schrödinger odd kitten states of $β=1.06$ and $β=1.11$ generated from one-photon-subtracted squeezed vacuum states of $-$3 dB, are amplified to an odd cat state of $β=1.73$ with a fidelity of $F=99\%$. A large-size Schrödinger odd cat state with $β=2.51$ and $F=97.30\%$ is predicted when the input squeezed vacuum states are increased to $-$5.91 dB. According to the analysis on the impacts of experimental imperfections in practice, Schrödinger odd cat states of $β>2$ are available. A feasible configuration based on a quantum frequency comb is developed to realize the large-size cat state generation scheme we proposed.

quant-ph↗

arXiv2021

Deep-CNN based Robotic Multi-Class Under-Canopy Weed Control in Precision Farming

Smart weeding systems to perform plant-specific operations can contribute to the sustainability of agriculture and the environment. Despite monumental advances in autonomous robotic technologies for precision weed management in recent years, work on under-canopy weeding in fields is yet to be realized. A prerequisite of such systems is reliable detection and classification of weeds to avoid mistakenly spraying and, thus, damaging the surrounding plants. Real-time multi-class weed identification enables species-specific treatment of weeds and significantly reduces the amount of herbicide use. Here, our first contribution is the first adequately large realistic image dataset \textit{AIWeeds} (one/multiple kinds of weeds in one image), a library of about 10,000 annotated images of flax, and the 14 most common weeds in fields and gardens taken from 20 different locations in North Dakota, California, and Central China. Second, we provide a full pipeline from model training with maximum efficiency to deploying the TensorRT-optimized model onto a single board computer. Based on \textit{AIWeeds} and the pipeline, we present a baseline for classification performance using five benchmark CNN models. Among them, MobileNetV2, with both the shortest inference time and lowest memory consumption, is the qualified candidate for real-time applications. Finally, we deploy MobileNetV2 onto our own compact autonomous robot \textit{SAMBot} for real-time weed detection. The 90\% test accuracy realized in previously unseen scenes in flax fields (with a row spacing of 0.2-0.3 m), with crops and weeds, distortion, blur, and shadows, is a milestone towards precision weed control in the real world. We have publicly released the dataset and code to generate the results at \url{https://github.com/StructuresComp/Multi-class-Weed-Classification}.

cs.CV↗

arXiv2021

LatentHuman: Shape-and-Pose Disentangled Latent Representation for Human Bodies

3D representation and reconstruction of human bodies have been studied for a long time in computer vision. Traditional methods rely mostly on parametric statistical linear models, limiting the space of possible bodies to linear combinations. It is only recently that some approaches try to leverage neural implicit representations for human body modeling, and while demonstrating impressive results, they are either limited by representation capability or not physically meaningful and controllable. In this work, we propose a novel neural implicit representation for the human body, which is fully differentiable and optimizable with disentangled shape and pose latent spaces. Contrary to prior work, our representation is designed based on the kinematic model, which makes the representation controllable for tasks like pose animation, while simultaneously allowing the optimization of shape and pose for tasks like 3D fitting and pose tracking. Our model can be trained and fine-tuned directly on non-watertight raw data with well-designed losses. Experiments demonstrate the improved 3D reconstruction performance over SoTA approaches and show the applicability of our method to shape interpolation, model fitting, pose tracking, and motion retargeting.

cs.CV↗

arXiv2021

Detecting quantum entanglement with unsupervised learning

Quantum properties, such as entanglement and coherence, are indispensable resources in various quantum information processing tasks. However, there still lacks an efficient and scalable way to detecting these useful features, especially for high-dimensional and multipartite quantum systems. In this work, we exploit the convexity of samples without the desired quantum features and design an unsupervised machine learning method to detect the presence of such features as anomalies. Particularly, in the context of entanglement detection, we propose a complex-valued neural network composed of pseudo-siamese network and generative adversarial net, and then train it with only separable states to construct non-linear witnesses for entanglement. It is shown via numerical examples, ranging from two-qubit to ten-qubit systems, that our network is able to achieve high detection accuracy which is above 97.5% on average.Moreover, it is capable of revealing rich structures of entanglement, such as partial entanglement among subsystems. Our results are readily applicable to the detection of other quantum resources such as Bell nonlocality and steerability, and thus our work could provide a powerful tool to extract quantum features hidden in multipartite quantum data.

quant-ph↗

arXiv2021

Small Signal Analysis of SQUID Direct Readout Schemes

To better understand the working principles of Superconducting Quantum Interference Device (SQUID) direct readout schemes, which work in different bias and amplifier modes with different internal feedback schemes, we present the complete circuit analyses on SQUID small-signal model. SQUID bias and amplifier circuits are analyzed using SQUID Thevenin equivalent circuit, and the general equivalent circuit of SQUID with different internal feedback schemes is derived and analyzed with a trans-impedance amplifier model. Transfer characteristics and noise performances of different direct readout schemes are analyzed and experimentally characterized. It is shown that amplifier noise suppression is only dependent on SQUID flux-to-voltage transfer coefficient and is irrelevant to the configuration of bias and amplifier; SQUID with an internal feedback scheme improves the transfer coefficient with voltage feedback and regulates the dynamic resistance with current feedback.

physics.ins-det↗

arXiv2021

Learning Object-Compositional Neural Radiance Field for Editable Scene Rendering

Implicit neural rendering techniques have shown promising results for novel view synthesis. However, existing methods usually encode the entire scene as a whole, which is generally not aware of the object identity and limits the ability to the high-level editing tasks such as moving or adding furniture. In this paper, we present a novel neural scene rendering system, which learns an object-compositional neural radiance field and produces realistic rendering with editing capability for a clustered and real-world scene. Specifically, we design a novel two-pathway architecture, in which the scene branch encodes the scene geometry and appearance, and the object branch encodes each standalone object conditioned on learnable object activation codes. To survive the training in heavily cluttered scenes, we propose a scene-guided training strategy to solve the 3D space ambiguity in the occluded regions and learn sharp boundaries for each object. Extensive experiments demonstrate that our system not only achieves competitive performance for static scene novel-view synthesis, but also produces realistic rendering for object-level editing.

cs.CV↗

arXiv2021

Control of continuous-mode single-photon states: a review

In this survey, we first introduce quantum fields and open quantum systems, then we present continuous-mode single-photon states and discuss discrete measurements of a single-photon field. After that, we introduce linear quantum systems and show how a linear quantum system responds to a single-photon input. Then we investigate how a coherent feedback network can be used to manipulate the temporal pulse shape of a single-photon state. Afterwards, we present single-photon filter and master equations. Finally, we discuss the generation of Schrödinger cat states by means of photon addition and subtraction

quant-ph↗

arXiv2021

Quantum compressed sensing of ultra-wideband radio-frequency signal

Real-time sensing of ultra-wideband radio-frequency signal with high frequency resolution is challenging, which is confined by the sampling rate of electronic analog-to-digital converter and the capability of digital signal processing. By combining quantum mechanics with compressed sensing, quantum compressed sensing is proposed for wideband radio-frequency signal frequency measurement. By using an electro-optical crystal as a sensor which modulates the wave function of the coherent photons with the signal to be measured. The frequency spectrum could be recovered by detecting the modulated sparse photons with a low time-jitter single-photon detector and a time-to-digital converter. More than 50 GHz real-time analysis bandwidth is demonstrated with the Fourier transform limit resolution. The further simulation shows it can be extended to more than 300 GHz with the present technologies.

quant-ph↗

arXiv2021

VS-Net: Voting with Segmentation for Visual Localization

Visual localization is of great importance in robotics and computer vision. Recently, scene coordinate regression based methods have shown good performance in visual localization in small static scenes. However, it still estimates camera poses from many inferior scene coordinates. To address this problem, we propose a novel visual localization framework that establishes 2D-to-3D correspondences between the query image and the 3D map with a series of learnable scene-specific landmarks. In the landmark generation stage, the 3D surfaces of the target scene are over-segmented into mosaic patches whose centers are regarded as the scene-specific landmarks. To robustly and accurately recover the scene-specific landmarks, we propose the Voting with Segmentation Network (VS-Net) to segment the pixels into different landmark patches with a segmentation branch and estimate the landmark locations within each patch with a landmark location voting branch. Since the number of landmarks in a scene may reach up to 5000, training a segmentation network with such a large number of classes is both computation and memory costly for the commonly used cross-entropy loss. We propose a novel prototype-based triplet loss with hard negative mining, which is able to train semantic segmentation networks with a large number of labels efficiently. Our proposed VS-Net is extensively tested on multiple public benchmarks and can outperform state-of-the-art visual localization methods. Code and models are available at \href{https://github.com/zju3dv/VS-Net}{https://github.com/zju3dv/VS-Net}.

cs.CV↗

arXiv2021

LIFE: Lighting Invariant Flow Estimation

We tackle the problem of estimating flow between two images with large lighting variations. Recent learning-based flow estimation frameworks have shown remarkable performance on image pairs with small displacement and constant illuminations, but cannot work well on cases with large viewpoint change and lighting variations because of the lack of pixel-wise flow annotations for such cases. We observe that via the Structure-from-Motion (SfM) techniques, one can easily estimate relative camera poses between image pairs with large viewpoint change and lighting variations. We propose a novel weakly supervised framework LIFE to train a neural network for estimating accurate lighting-invariant flows between image pairs. Sparse correspondences are conventionally established via feature matching with descriptors encoding local image contents. However, local image contents are inevitably ambiguous and error-prone during the cross-image feature matching process, which hinders downstream tasks. We propose to guide feature matching with the flows predicted by LIFE, which addresses the ambiguous matching by utilizing abundant context information in the image pairs. We show that LIFE outperforms previous flow learning frameworks by large margins in challenging scenarios, consistently improves feature matching, and benefits downstream tasks.

cs.CV↗

arXiv2020

Super interference fringes of two-photon photoluminescence in individual Au nanoparticles: the critical role of the intermediate state

The interaction between light and metal nanoparticles enables investigations of microscopic phenomena on nanometer length and ultrashort time scales, benefiting from strong confinement and enhancement of the optical field. However, the ultrafast dynamics of these nanoparticles are primarily investigated by multiphoton photoluminescence on picoseconds or photoemission on femtoseconds independently. Here, we presented two-photon photoluminescence (TPPL) measurements on individual Au nanobipyramids (AuNP) to reveal their ultrafast dynamics by two-pulse excitation on a global time scale ranging from sub-femtosecond to tens of picoseconds. Two-orders-of-magnitude photoluminescence enhancement, namely super interference fringes, has been demonstrated on tens of femtoseconds. Power-dependent measurements uncovered the transform of the nonlinearity from 1 to 2 when the interpulse delay varied from tens of femtoseconds to tens of picoseconds. We proved that the real intermediate state plays a critical role in the observed phenomena, supported by numerical simulations with a three eigenstates model and further experiments on Au nanospheres with different diameters. The crucial parameters, including the dephasing time, the radiative rate, and the coupling between different states, have been estimated using numerical simulations. Our results provide insight into the role of intermediate states in the ultrafast dynamics of noble metal nanoparticles. The giant photoluminescence in super interference fringes enables potential practical applications in imaging, sensing, and nanophotonics.

physics.optics↗

arXiv2020

Multi-observable Uncertainty Equality based on the sum of standard deviations in the qubit system

We construct a multi-observable uncertainty equality as well as an inequality based on the sum of standard deviations in the qubit system. The obtained equality indicates that the uncertainty relation can be expressed more accurately, and also can be used to detect the mixedness of the system. Meanwhile, the new uncertainty inequality can provide a tighter lower bound, and the tightness can be maintained at a high level even in an open system. Furthermore, the deficiency of the uncertainty relation, that the lower bound of the product form uncertainty relations can be null even for two incompatible observables, can be completely fixed by the new uncertainty relation.

quant-ph↗

arXiv2020

On the dynamics of a quantum coherent feedback network of cavity-mediated double quantum dot qubits

The purpose of this paper is to present a comprehensive study of a coherent feedback network where the main component consists of two distant double quantum dot (DQD) qubits which are directly coupled to a cavity. This main component has recently been physically realized (van Woerkom, {\it et al.}, Microwave photon-mediated interactions between semiconductor qubits, Physical Review X, 8(4):041018, 2018). The feedback loop is closed by cascading this main component with a beamsplitter. The dynamics of this coherent feedback network is studied from three perspectives. First, an analytic form of the output single-photon state of the network driven by a single-photon state is derived; in particular, it is observed that coherent feedback elongates considerably the interaction between the input single photon and the network. Second, excitation probabilities of DQD qubits are computed when the network is driven by a single-photon input state. Moreover, if the input is vacuum but one of the two DQD qubits is initialized in its excited state, the explicit expression of the state of the network is derived, in particular, it is shown that the output field and the two DQD qubits can form an entangled state if the transition frequencies of two DQD qubits are equal. Finally, the exact form of the pulse shape is obtained by which the single-photon input can fully excite one of these two DQD qubits at any controllable time, which may be useful in the construction of $2$-qubit quantum gates.

quant-ph↗

arXiv2020

Quantum Higher Order Singular Value Decomposition

Higher order singular value decomposition (HOSVD) is an important tool for analyzing big data in multilinear algebra and machine learning. In this paper, we present two quantum algorithms for HOSVD. Our methods allow one to decompose a tensor into a core tensor containing tensor singular values and some unitary matrices by quantum computers. Compared to the classical HOSVD algorithm, our quantum algorithms provide an exponential speedup. Furthermore, we introduce a hybrid quantum-classical algorithm of HOSVD model applied in recommendation systems.

quant-ph↗

arXiv2020

On the dynamics of two photons interacting with a two-qubit coherent feedback network}

The purpose of this paper is to study the dynamics of a quantum coherent feedback network composed of two two-level systems (qubits) driven by two counter-propagating photons, one in each input channel. The coherent feedback network enhances the nonlinear photon-photon interaction inside the feedback loop. By means of quantum stochastic calculus and the input-output framework, the analytic form of the steady-state output two-photon state is derived. Based on the analytic form, the applications on the Hong-Ou-Mandel (HOM) interferometer and marginally stable single-photon devices using this coherent feedback structure have been demonstrated. The difference between continuous-mode and single-mode few-photon states is demonstrated.

quant-ph↗