- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
[IEEE 2018 24th International Conference on Pattern Recognition (ICPR) - Beijing, China (2018.8.20-2018.8.24)] 2018 24th International Conference on Pattern Recognition (ICPR) - RelationNet: Learning Deep-Aligned Representation for Semantic Image Segmentation
摘要： Semantic image segmentation, which assigns labels in pixel level, plays a central role in image understanding. Recent approaches have attempted to harness the capabilities of deep learning. However, one central problem of these methods is that deep convolutional neural network gives little consideration to the correlation among pixels. To handle this issue, in this paper, we propose a novel deep neural network named RelationNet, which utilizes CNN and RNN to aggregate context information. Besides, a spatial correlation loss is applied to train RelationNet to align features of spatial pixels belonging to same category. Importantly, since it is expensive to obtain pixel-wise annotations, we exploit a new training method to combine the coarsely and finely labeled data. Experiments show the detailed improvements of each proposal. Experimental results demonstrate the effectiveness of our proposed method to the problem of semantic image segmentation, which obtains state-of-the-art performance on the Cityscapes benchmark and Pascal Context dataset.
关键词： Spatial correlation loss,CNN,Semantic image segmentation,RNN,Deep learning,RelationNet
更新于2025-09-04 15:30:14
-
[IEEE 2018 24th International Conference on Pattern Recognition (ICPR) - Beijing, China (2018.8.20-2018.8.24)] 2018 24th International Conference on Pattern Recognition (ICPR) - Large Margin Structured Convolution Operator for Thermal Infrared Object Tracking
摘要： Compared with visible object tracking, thermal infrared (TIR) object tracking can track an arbitrary target in total darkness since it cannot be influenced by illumination variations. However, there are many unwanted attributes that constrain the potentials of TIR tracking, such as the absence of visual color patterns and low resolutions. Recently, structured output support vector machine (SOSVM) and discriminative correlation filter (DCF) have been successfully applied to visible object tracking, respectively. Motivated by these, in this paper, we propose a large margin structured convolution operator (LM-SCO) to achieve efficient TIR object tracking. To improve the tracking performance, we employ the spatial regularization and implicit interpolation to obtain continuous deep feature maps, including deep appearance features and deep motion features, of the TIR targets. Finally, a collaborative optimization strategy is exploited to significantly update the operators. Our approach not only inherits the advantage of the strong discriminative capability of SOSVM but also achieves accurate and robust tracking with higher-dimensional features and more dense samples. To the best of our knowledge, we are the first to incorporate the advantages of DCF and SOSVM for TIR object tracking. Comprehensive evaluations on two thermal infrared tracking benchmarks, i.e. VOT-TIR2015 and VOT-TIR2016, clearly demonstrate that our LMSCO tracker achieves impressive results and outperforms most state-of-the-art trackers in terms of accuracy and robustness with sufficient frame rate.
关键词： thermal infrared object tracking,deep motion features,deep appearance features,large margin structured convolution operator,structured output support vector machine,discriminative correlation filter
更新于2025-09-04 15:30:14
-
[IEEE 2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI) - Las Vegas, NV (2018.4.8-2018.4.10)] 2018 IEEE Southwest Symposium on Image Analysis and Interpretation (SSIAI) - In-between and cross-frequency dependence-based summarization of resting-state fMRI data
摘要： Various data summarization approaches which consist of basis transformation and dimension reduction have been commonly used for information retrieval from brain imaging data including functional magnetic resonance imaging (fMRI). However, most approaches do not include frequency variation of the temporal data in the basis transformation. Here we propose a novel approach to incorporate in-between and cross-frequency dependence for summarization of resting-state fMRI data.
关键词： Cross-frequency dependence,Hilbert transform,Canonical correlation analysis
更新于2025-09-04 15:30:14
-
Digital Image Correlation for discontinuous displacement measurement using subset segmentation
摘要： Deformation measurement is normally achieved by using Digital Image Correlation (DIC) technique when deformation is not discontinuous. However, the presence of discontinuities makes the deformation process very challenging and DIC fails. An innovative technique is proposed in this study which splits the subset (segment) of an image into multiple parts and use segmented subset of the image for correlation process. The performance of the proposed technique is evaluated using di?erent experiments where di?erent types of discontinuities are introduced in the deformation process at di?erent angles and having di?erent discontinuity opening sizes. The obtained results are compared with the recently proposed Discontinuous Digital Image Correlation (DDIC) technique. The results show that the proposed technique is more reliable and having high accuracy which reaches upto 1/100th of a pixel under favorable circumstances.
关键词： Deformation measurement,Discontinuous displacement measurement,Digital image correlation,Reconstruction of displacement ?elds,DIC,Subset segmentation
更新于2025-09-04 15:30:14
-
Entanglement, nonclassical properties, and geometric phase in circuit quantum electrodynamics with relativistic motion
摘要： In this work, we study qualitatively the nonclassical properties, nonlocal correlation, and geometric phase in a superconducting qubit system connected to a transmission line resonator (TLR) within a circuit quantum electrodynamics setup, where the modulation of the coupling strength can imitate the harmonic motion of the superconducting system at relativistic speeds. We analyze the effects of the relativistic motion on these quantum quantifiers under the influence of the physical parameters implicated in the whole quantum system state. Furthermore, the relationship between the different quantifiers is explored. Our results may open new perspectives to understand and explore the relativistic phenomena on the comportment of the quantum quantifiers for new applications in superconducting quantum technologies.
关键词： E. Nonclassical properties and correlation,B. Superconducting qubit system,A. Solid state systems,C. Transmission line resonator,D. Circuit quantum electrodynamics
更新于2025-09-04 15:30:14
-
Compositional correlation and polymorphism in BaF2-PrF3 thin films deposited using electron-beam evaporation
摘要： Infrared low-index evaporation materials are essential in the broadband antire?ection coatings to reduce Fresnel re?ections on the surfaces of infrared optical components. Although the layers of praseodymium ?uoride (PrF3) show an excellent transparency and the lower refractive index n and extinction coe?cient k in the spectral range of thermal infrared, the tensile stress presented in the layers prohibits PrF3 from being used as infrared low-index coating materials. A practical solution to reduce stress is to directly evaporate the admixture of PrF3 with alkaline ?uorides, such as barium ?uoride (BaF2). However, due to the signi?cant di?erence of vapor pressures between PrF3 and BaF2, it is commonly di?cult to congruently deposit PrF3-BaF2 thin ?lms utilizing evaporating directly from a single source. Moreover, more details are unknown about the phase compositions of PrF3-BaF2 thin ?lms. In our investigation, BaF2-PrF3 thin ?lms were deposited using electron beam evaporation from the sintered ingots of PrF3 admixed with BaF2. The compositions of thin ?lms were characterized using energy dispersive X-ray spectroscopy (EDX), the crystallographic structures were explored by X-ray-di?raction (XRD). It was revealed that the concentration of BaF2 in thin ?lms can be correlated to that in the admixtures. Moreover, in addition to PrF3 and BaF2, the phase compositions in thin ?lms include also secondary phases, such as PrOF, PrF4, Pr2F2, BaPrF6, and that of elemental Ba.
关键词： Infrared low-index evaporation materials,BaF2-PrF3,Compositional correlation,Polymorphism,Electron beam evaporation
更新于2025-09-04 15:30:14
-
Quantum imaging with sub-Poissonian light: challenges and perspectives in optical metrology
摘要： Non-classical correlations in optical beams offer the unprecedented opportunity of surpassing conventional limits of sensitivity and resolution in optical measurements and imaging, especially but not only, when a low photon flux down/single photons are measured. We review the principles of quantum imaging and sensing techniques which exploit sub-Poissonian photon statistics and non-classical photon number correlation, presenting some state-of-the-art achievements in the field. These quantum photonics protocols have the potential to trigger major steps in many applications, such as microscopy and biophotonics and represent an important opportunity for a new deal in radiometry and photometry.
关键词： quantum optics,quantum correlation,sub-Poissonian light,Quantum enhanced measurement
更新于2025-09-04 15:30:14
-
Extracting Displacement and Strain Fields from Checkerboard Images with the Localized Spectrum Analysis
摘要： The performance of white-light full-field measurement methods strongly depends on the nature of the pattern used to mark the surface on which displacements and strains are measured. Finding optimized patterns is therefore a topical question. The aim of this study is to examine the case of the checkerboard pattern. It is first shown that this periodic pattern can be processed with a Fourier-based technique such as LSA. Experiments are then carried out to compare the noise level in displacement and strain maps obtained by processing classic 2D grid and checkerboard images. The conclusion is that the noise level observed in displacement and strain maps is significantly lower with a checkerboard than with a classic 2D grid. A notched specimen is finally tested to illustrate that very low strain levels can be measured with checkerboard patterns.
关键词： Grid method,Checkerboard,Localized spectrum analysis,Optimal pattern,Digital image correlation,Heteroscedastic noise,Metrology,Pattern optimization,Windowed Fourier transform
更新于2025-09-04 15:30:14
-
[IEEE 2018 15th European Radar Conference (EuRAD) - Madrid, Spain (2018.9.26-2018.9.28)] 2018 15th European Radar Conference (EuRAD) - Robust Correlation based True-Speed-over-Ground Measurement System employing a FMCW Radar
摘要： This paper presents a novel approach for the determination of True-Speed-Over-Ground for trains. Speed determination is accomplished by correlating the received signals of two side-looking radar sensors. The theoretically achievable precision is derived. Test measurements are done to give a prove of concept. The measurements are done in two different scenarios. The test measurements are carried out using a bi-static 24 GHz frequency modulated continuous wave (FMCW) radar.
关键词： radar,correlation,fmcw,velocity,sar,train
更新于2025-09-04 15:30:14
-
[IEEE 2017 International Conference on Optical Network Design and Modeling (ONDM) - Budapest (2017.5.15-2017.5.18)] 2017 International Conference on Optical Network Design and Modeling (ONDM) - A probabilistic approach for failure localization
摘要： This work considers the problem of fault localization in transparent optical networks. The aim is to localize single-link failures by utilizing statistical machine learning techniques trained on data that describe the network state upon current and past failure incidents. In particular, a Gaussian Process (GP) classi?er is trained on historical data extracted from the examined network, with the goal of modeling and predicting the failure probability of each link therein. To limit the set of suspect links for every failure incident, the proposed approach is complemented with the utilization of a Graph-Based Correlation heuristic. The proposed approach is tested on a dataset generated for an OFDM-based optical network, demonstrating that it achieves a high localization accuracy. The proposed scheme can be used by service providers for reducing the Mean-Time-To-Repair of the failure.
关键词： Graph-Based Correlation,OFDM,Gaussian Process,transparent optical networks,fault localization
更新于2025-09-04 15:30:14