- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Revealing the Spectrum of Unknown Layered Materials with Super-Human Predictive Abilities
摘要： We discover the chemical composition of over 1000 materials that are likely to exhibit layered and two-dimensional phases but have yet to be synthesized. This includes two materials our calculations indicate can exist in distinct structures with different band gaps, expanding the short list of two-dimensional phase change materials. While databases of over 1000 layered materials have been reported, we provide the first full database of materials that are likely layered but yet to be synthesized, providing a roadmap for the synthesis community. We accomplish this by combining physics with machine learning on experimentally obtained data and verify a subset of candidates using density functional theory. We find our model performs five times better than practitioners in the field at identifying layered materials and is comparable or better than professional solid-state chemists. Finally, we find that semi-supervised learning can offer benefits for materials design where labels for some of the materials are unknown.
关键词： two-dimensional materials,machine learning,materials discovery,density functional theory,layered materials
更新于2025-09-09 09:28:46
-
European Microscopy Congress 2016: Proceedings || A new method for quantitative XEDS tomography of complex hetero-nanostructures
摘要： Over the last decades, electron tomography (ET) of complex materials has evolved into a powerful tool to investigate the three-dimensional (3D) structure of materials at the nanometer scale. The technique is based on the acquisition of a series of two-dimensional (2D) projection images of a sample, which are then reconstructed into a 3D volume using computational methods. Despite its success, ET faces several challenges, including the missing wedge problem, which limits the resolution and fidelity of the reconstructed volume. In this paper, we present a novel approach to mitigate the missing wedge problem by combining ET with machine learning (ML). Our method leverages the ability of ML to learn from data and predict missing information, thereby improving the quality of the reconstructed volume. We demonstrate the effectiveness of our approach on a series of synthetic and experimental datasets, showing significant improvements in resolution and fidelity compared to traditional reconstruction methods.
关键词： machine learning,nanomaterials,3D reconstruction,missing wedge,electron tomography
更新于2025-09-09 09:28:46
-
[SPIE Biomedical Applications in Molecular, Structural, and Functional Imaging - Houston, United States (2018.2.10-2018.2.15)] Medical Imaging 2018: Biomedical Applications in Molecular, Structural, and Functional Imaging - Automatic quantification framework to detect cracks in teeth
摘要： Studies show that cracked teeth are the third most common cause for tooth loss in industrialized countries. If detected early and accurately, patients can retain their teeth for a longer time. Most cracks are not detected early because of the discontinuous symptoms and lack of good diagnostic tools. Currently used imaging modalities like Cone Beam Computed Tomography (CBCT) and intraoral radiography often have low sensitivity and do not show cracks clearly. This paper introduces a novel method that can detect, quantify, and localize cracks automatically in high resolution CBCT (hr-CBCT) scans of teeth using steerable wavelets and learning methods. These initial results were created using hr-CBCT scans of a set of healthy teeth and of teeth with simulated longitudinal cracks. The cracks were simulated using multiple orientations. The crack detection was trained on the most significant wavelet coefficients at each scale using a bagged classifier of Support Vector Machines. Our results show high discriminative specificity and sensitivity of this method. The framework aims to be automatic, reproducible, and open-source. Future work will focus on the clinical validation of the proposed techniques on different types of cracks ex-vivo. We believe that this work will ultimately lead to improved tracking and detection of cracks allowing for longer lasting healthy teeth.
关键词： High-resolution Cone Beam Computed Tomography,Machine learning,Wavelet analysis,Tooth fracture detection
更新于2025-09-09 09:28:46
-
[ACM Press SIGGRAPH Asia 2018 Technical Briefs - Tokyo, Japan (2018.12.04-2018.12.07)] SIGGRAPH Asia 2018 Technical Briefs on - SA '18 - Learning photo enhancement by black-box model optimization data generation
摘要： We address the problem of automatic photo enhancement, in which the challenge is to determine the optimal enhancement for a given photo according to its content. For this purpose, we train a convolutional neural network to predict the best enhancement for given picture. While such machine learning techniques have shown great promise in photo enhancement, there are some limitations. One is the problem of interpretability, i.e., that it is not easy for the user to discern what has been done by a machine. In this work, we leverage existing manual photo enhancement tools as a black-box model, and predict the enhancement parameters of that model. Because the tools are designed for human use, the resulting parameters can be interpreted by their users. Another problem is the di?culty of obtaining training data. We propose generating supervised training data from high-quality professional images by randomly sampling realistic de-enhancement parameters. We show that this approach allows automatic enhancement of photographs without the need for large manually labelled supervised training datasets.
关键词： machine learning,photo enhancement,black-box optimization
更新于2025-09-09 09:28:46
-
[IEEE 2018 2nd International Conference On Electrical Engineering (EECON) - Colombo, Sri Lanka (2018.9.28-2018.9.28)] 2018 2nd International Conference On Electrical Engineering (EECon) - Application of Machine Learning Algorithms for Solar Power Forecasting in Sri Lanka
摘要： Reliability and stability of a power system get decrease with the integration of large proportion of renewable energy. Renewable sources such as solar and wind are highly intermittent, and it is difficult to maintain system stability with intolerable proportion of renewable energy injection. Solar power forecasting can be used to improve system stability by providing approximated future power generation to system control engineers and it will facilitate dispatch of hydro power plants in an optimum way. Machine Learning (ML) algorithms have shown great performance in time series forecasting and hence can be used to forecast power using weather parameters as model inputs. This paper presents the application of several ML algorithms for solar power forecasting in Buruthakanda solar park situated in Hambantota, Sri Lanka. The forecasting performance of implemented ML algorithms is compared with Smart Persistence (SP) method and the research shows that the ML models outperforms SP model.
关键词： solar power forecasting,solar power in Sri Lanka,machine learning for forecasting,renewable energy
更新于2025-09-09 09:28:46
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Deep Learning Neural Networks for Land Use Land Cover Mapping
摘要： The importance of accurate and timely information describing the nature and extent of land resources and changes over time is increasing. This research examines the application of deep learning neural networks (DLNN) to the analysis of satellite imagery with specific focus on the production of land use/land cover maps. DLNN have made considerable strides in pattern recognition and machine learning over the last several years. However, their application to remote sensing is less well developed as the technology was originally designed for simple photographs and not satellite imagery. This research presents the results of an experimental study conducted that developed a DLNN to generate land use/land cover maps of the southern agricultural region of Manitoba, Canada. The results of this approach demonstrate a clear advantage in processing time once the DLNN is properly trained when compared to human based semi-automated process.
关键词： Neural Networks,Land Cover Mapping,Big Data,Machine Learning
更新于2025-09-09 09:28:46
-
Machine Learning for Perovskites' Reap-Rest-Recovery Cycle
摘要： Perovskite photovoltaics are efficient and inexpensive, yet their performance is dynamic. In this Perspective, we examine the effects of H2O, O2, bias, temperature, and illumination on device performance and recovery. First, we discuss pivotal experiments that evaluate perovskites’ ability to go through a reap-rest-recovery (3R) cycle, and how machine learning (ML) can help identify the optimum values for each operating parameter. Second, we analyze perovskite dynamics and degradation, emphasizing the research challenges surrounding this 3R cycle. We then outline experiments that could identify the impact of environmental factors on recovery for different perovskite compositions. Finally, we propose an ML paradigm for maximizing long-term performance and predicting device performance recovery, including a shared-knowledge repository. By reframing perovskites’ optoelectronic transiency within the context of recovery rather than degradation, we highlight a set of research opportunities and the artificial intelligence solutions needed for the commercial adoption of these promising solar cell materials.
关键词： machine learning,perovskite photovoltaics,environmental factors,reap-rest-recovery cycle,device performance
更新于2025-09-09 09:28:46
-
Soft optoelectronic sensory foams with proprioception
摘要： In a step toward soft robot proprioception, and therefore better control, this paper presents an internally illuminated elastomer foam that has been trained to detect its own deformation through machine learning techniques. Optical fibers transmitted light into the foam and simultaneously received diffuse waves from internal reflection. The diffuse reflected light was interpreted by machine learning techniques to predict whether the foam was twisted clockwise, twisted counterclockwise, bent up, or bent down. Machine learning techniques were also used to predict the magnitude of the deformation type. On new data points, the model predicted the type of deformation with 100% accuracy and the magnitude of the deformation with a mean absolute error of 0.06°. This capability may impart soft robots with more complete proprioception, enabling them to be reliably controlled and responsive to external stimuli.
关键词： machine learning,elastomer foam,soft robots,optical fibers,proprioception
更新于2025-09-09 09:28:46
-
Blind Image Quality Assessment Using Multiscale Local Binary Patterns
摘要： This article proposes a new no-reference image quality assessment method that is able to blindly predict the quality of an image. The method is based on a machine learning technique that uses texture descriptors. In the proposed method, texture features are computed by decomposing images into texture information using multiscale local binary pattern (MLBP) operators. In particular, the parameters of local binary pattern operators are varied, which generates MLBP operators. The features used for training the prediction algorithm are the histograms of these MLBP channels. The results show that, when compared with other state-of-the-art no-reference methods, the proposed method is competitive in terms of prediction precision and computational complexity.
关键词： MLBP,machine learning,multiscale local binary pattern,texture descriptors,no-reference image quality assessment
更新于2025-09-09 09:28:46
-
[IEEE 2018 15th European Radar Conference (EuRAD) - Madrid, Spain (2018.9.26-2018.9.28)] 2018 15th European Radar Conference (EuRAD) - Image-Based Pedestrian Classification for 79 GHz Automotive Radar
摘要： Radar sensors have become an integral part of advanced driver assistance systems. Merely detecting targets will not, however, advance their contribution. Rather, an object classification capability is required to distinguish vulnerable road users from other objects, such as vehicles. In this paper, we present a novel pedestrian classification procedure, which uses image features from the range-Doppler-Matrix created by a 79 GHz chirp sequence radar. Experiments show single measurement success rates of 88% for a bandwidth of 1.6 GHz. Moreover, the robustness of the classification process is consolidated with a tracking algorithm. Implemented in vehicles, this can be a major contribution to protect vulnerable road users.
关键词： machine learning,79 GHz,chirp sequence,automotive radar,pedestrian classification,SURF
更新于2025-09-09 09:28:46