- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
Dynamic spectrum matching with one-shot learning
摘要： Convolutional neural networks (CNN) have been shown to provide a good solution for classification problems that utilize data obtained from vibrational spectroscopy. Moreover, CNNs are capable of identifying substances from noisy spectra without the need for additional preprocessing. However, their application in practical spectroscopy is restricted due to two reasons. First the effectiveness of classification using CNNs diminishes rapidly when only a small number of spectra per substance are available for training (which is a typical situation in real applications). Secondly, to accommodate new, previously unseen, substance classes the network must be retrained which is computationally intensive. Here we address these issues by reformulating a multi-class classification problem with a large number of classes to a binary classification problem for which the available data is sufficient for representation learning. Hence, we define the learning task as identifying pairs of inputs as belonging to the same class or different classes. We achieve this using a Siamese convolutional neural network. A novel sampling strategy is proposed to address the imbalance problem in training the Siamese network. The trained network can classify samples of previously unseen substance classes using just a single reference sample (termed as one-shot learning in the machine learning community). Our results on three independent Raman datasets demonstrate much better accuracy than other practical systems to date, while allowing effortless updates of the system’s database with new substance classes.
关键词： Spectrum Matching,Siamese Network,Convolutional Neural Networks,One-shot Learning
更新于2025-09-23 15:19:57
-
[IEEE 2018 26th European Signal Processing Conference (EUSIPCO) - Roma, Italy (2018.9.3-2018.9.7)] 2018 26th European Signal Processing Conference (EUSIPCO) - END-to-END Photopleth YsmographY (PPG) Based Biometric Authentication by Using Convolutional Neural Networks
摘要： Whilst research efforts have traditionally focused on Electrocardiographic (ECG) signals and handcrafted features as potential biometric traits, few works have explored systems based on the raw photoplethysmogram (PPG) signal. This work proposes an end-to-end architecture to offer biometric authentication using PPG biosensors through Convolutional Networks. We provide an evaluation of the performance of our approach in two different databases: Troika and PulseID, the latter a publicly available database specifically collected by the authors for such a purpose. Our verification approach through convolutional network based models and using raw PPG signals appears to be viable in current monitoring procedures within e-health and fitness environments showing a remarkable potential as a biometry. The approach tested on a verification fashion, on trials lasting one second, achieved an AUC of 78.2% and 83.2%, averaged among target subjects, on PulseID and Troika datasets respectively. Our experimental results on previous small datasets support the usefulness of PPG extracted biomarkers as viable traits for multi-biometric or standalone biometrics. Furthermore, the approach results in a low input throughput and complexity that allows for a continuous authentication in real-world scenarios. Nevertheless, the reported experiments also suggest that further research is necessary to account for and understand sources of variability found in some subjects.
关键词： biometric verification,photoplethysmogram signal,convolutional neural networks,biometric authentication,ppg
更新于2025-09-23 15:19:57
-
[IEEE 2018 IEEE International Conference on Cyborg and Bionic Systems (CBS) - Shenzhen, China (2018.10.25-2018.10.27)] 2018 IEEE International Conference on Cyborg and Bionic Systems (CBS) - RDS-Denoiser: a Detail-preserving Convolutional Neural Network for Image Denoising
摘要： Image noise is usually modeled as additive independent Gaussian random variables with fixed standard deviation, and most existing methods developed under this assumption have difficulties handling spatially varying noise. In this work, we aim to solve the problem of image denoising when the noise level is unknown. We propose a simple yet effective Stacked Denoising Networks. It decomposes the denoising process into two stages. The Stage-I Denoising is to predict the noise map of noisy image. The Stage-II Denoising to further improve the visual quality and alleviate overfitting to Gaussian noise. Experiments show that RDS-Denoiser achieves competitive performance comparing to state-of-the-art denoising methods. In addition, we propose RDS-GAN, a conditional generative adversarial network, to further improve the visual quality and alleviate overfitting to Gaussian noise.
关键词： Image Denoising,Convolutional Neural Networks,Conditional Generative Adversarial Networks
更新于2025-09-19 17:15:36
-
[IEEE 2018 Digital Image Computing: Techniques and Applications (DICTA) - Canberra, Australia (2018.12.10-2018.12.13)] 2018 Digital Image Computing: Techniques and Applications (DICTA) - Robust CNN-based Gait Verification and Identification using Skeleton Gait Energy Image
摘要： As a kind of behavioral biometric feature, gait has been widely applied for human verification and identification. Approaches to gait recognition can be classified into two categories: model-free approaches and model-based approaches. Model-free approaches are sensitive to appearance changes. For model-based approaches, it is difficult to extract the reliable body models from gait sequences. In this paper, based on the robust skeleton points produced from a two-branch multi-stage CNN network, a novel model-based feature, Skeleton Gait Energy Image (SGEI), has been proposed. Relevant experimental performances indicate that SGEI is more robust to the cloth changes. Another contribution is that two different CNN-based architectures have been separately proposed for gait verification and gait identification. Both these two architectures have been evaluated on the datasets. They have presented satisfying performances and increased the robustness for gait recognition in the unconstrained environments with view variances and cloth variances.
关键词： Gait Identification,Gait Verification,Deep Convolutional Neural Networks,Skeleton Gait Energy Image
更新于2025-09-19 17:15:36
-
[IEEE 2018 25th IEEE International Conference on Image Processing (ICIP) - Athens (2018.10.7-2018.10.10)] 2018 25th IEEE International Conference on Image Processing (ICIP) - Curvature Augmented Deep Learning for 3D Object Recognition
摘要： This paper presents a new method to incorporate shape information into convolutional neural network (CNN)s for 3D object recognition. Voxel CNNs have been very successful with the task of 3D object recognition. However, continuous shape information that is useful for recognition is often lost in their conversion to a voxel representation. We propose a single dimensional feature that can be applied to voxel CNNs. This paper presents a novel rotation-invariant feature based on mean curvature that improves shape recognition for voxel CNNs. We augment the recent voxel CNN Octnet architecture with our feature and demonstrate a 1% overall accuracy increase on the ModelNet10 dataset.
关键词： 3D Object Recognition,Convolutional Neural Networks,Computational Geometry,Deep Learning
更新于2025-09-19 17:15:36
-
[IEEE 2018 International Conference on Information and Communication Technology Convergence (ICTC) - Jeju (2018.10.17-2018.10.19)] 2018 International Conference on Information and Communication Technology Convergence (ICTC) - Deep Inverse Tone Mapping Optimized for High Dynamic Range Display
摘要： The popularity of high dynamic range (HDR) makes the inverse tone mapping become an important technique for HDR display. In this paper, we propose a convolutional neural network (CNN) based inverse tone mapping method to generate a high-quality HDR image from one single standard dynamic range (SDR) image. First, we present a CNN design with a three-channel input, which considers both luminance and chrominance. Second, we propose to use overlapped inputs to remove the boundary artifacts, caused by zero padding in CNN. Experimental results demonstrate the high quality of our generated HDR images compared to the ground truth and conventional inverse tone mapping methods.
关键词： inverse tone mapping,high dynamic range imaging,convolutional neural networks
更新于2025-09-19 17:15:36
-
[IEEE 2018 International Joint Conference on Neural Networks (IJCNN) - Rio de Janeiro (2018.7.8-2018.7.13)] 2018 International Joint Conference on Neural Networks (IJCNN) - Defect classification in shearography images using convolutional neural networks
摘要： High subjectivity, lack of attention and fatigue are factors inherent to human analysis in inspection activities such as shearography, a non-destructive optical method. In order to minimize the probability of human error, a study was conducted in which a binary classification from 256 shearography test samples obtained from pipes repaired with glass fiber patches was performed. The dataset was split into major and minor defects and used to train two convolutional neural networks architectures, - a specific artificial neural network well known for its application on image classification. Architecture A achieved a maximum accuracy of 73% on major defect detection, while architecture B, slightly more complex, led to better results. Posterior studies on architecture B led to the conclusion that a combination of double layer filters and dropout layers are the best setup for this type of classification problem. It is possible that other architectures might lead to better results, but no grid search was performed to confirm this assumption. An accuracy of 79% was achieved with Architecture B, therefore is reasonable to say that convolutional neural networks are able to learn from parameters which are difficult to correctly process, such as the fringe patterns obtained from shearography test samples.
关键词： non-destructive testing,convolutional neural networks,composite material,shearography,binary classification
更新于2025-09-19 17:15:36
-
[IEEE 2018 IEEE 10th Latin-American Conference on Communications (LATINCOM) - Guadalajara, Jalisco, Mexico (2018.11.14-2018.11.16)] 2018 IEEE 10th Latin-American Conference on Communications (LATINCOM) - Convolutional Neural Networks for Semantic Segmentation of Multispectral Remote Sensing Images
摘要： The recent impulse in development of artificial intelligence (AI) methodologies has simplified the application of this in multiple research areas. This simplification was not favorable before, due to the limitations in dimensionality, processing time, computational resources, among others. Working with multispectral remote sensing (RS) images, in an artificial neural network (NN) was quite complex. Due the methods used required millions of processes that took a long time to be executed and produce competitive results compared with the state of the art (SoA). Deep learning (DL) strategies have been applied to alleviate these limitations and have greatly improved the use of neural networks. Therefore, this paper presents the analysis of DL-NNs to perform semantic segmentation of multispectral RS images. Images are captured by the constellation of satellites Sentinel-2 from the European Space Agency. The objective of this research is to classify each pixel of a scene into five categories: 1-vegetation, 2-soil, 3-water, 4-clouds and 5-cloud shadows. The selection of spectral bands for the formation of input datasets for segmentation of these classes is very important. The spectral signatures of each material aid to discern among several classes. Results presented in this work, show that the AI strategy proposed offer better accuracy segmentation than other methods of the SoA in competitive processing time.
关键词： semantic segmentation,Convolutional neural networks,remote sensing,multispectral images
更新于2025-09-19 17:15:36
-
Light Field Super-Resolution using a Low-Rank Prior and Deep Convolutional Neural Networks
摘要： Light ?eld imaging has recently known a regain of interest due to the availability of practical light ?eld capturing systems that offer a wide range of applications in the ?eld of computer vision. However, capturing high-resolution light ?elds remains technologically challenging since the increase in angular resolution is often accompanied by a signi?cant reduction in spatial resolution. This paper describes a learning-based spatial light ?eld super-resolution method that allows the restoration of the entire light ?eld with consistency across all angular views. The algorithm ?rst uses optical ?ow to align the light ?eld and then reduces its angular dimension using low-rank approximation. We then consider the linearly independent columns of the resulting low-rank model as an embedding, which is restored using a deep convolutional neural network (DCNN). The super-resolved embedding is then used to reconstruct the remaining views. The original disparities are restored using inverse warping where missing pixels are approximated using a novel light ?eld inpainting algorithm. Experimental results show that the proposed method outperforms existing light ?eld super-resolution algorithms, achieving PSNR gains of 0.23 dB over the second best performing method. The performance is shown to be further improved using iterative back-projection as a post-processing step.
关键词： Low-Rank Matrix Approximation,Super-Resolution,Light Field,Deep Convolutional Neural Networks
更新于2025-09-19 17:15:36
-
[SPIE Computer-Aided Diagnosis - Houston, United States (2018.2.10-2018.2.15)] Medical Imaging 2018: Computer-Aided Diagnosis - Convolutional neural networks for the detection of diseased hearts using CT images and left atrium patches
摘要： Cardiovascular disease is a leading cause of death in the United States. The identification of cardiac diseases on conventional three-dimensional (3D) CT can have many clinical applications. An automated method that can distinguish between healthy and diseased hearts could improve diagnostic speed and accuracy when the only modality available is conventional 3D CT. In this work, we proposed and implemented convolutional neural networks (CNNs) to identify diseased hearts on CT images. Six patients with healthy hearts and six with previous cardiovascular disease events received chest CT. After the left atrium for each heart was segmented, 2D and 3D patches were created. A subset of the patches were then used to train separate convolutional neural networks using leave-one-out cross-validation of patient pairs. The results of the two neural networks were compared, with 3D patches producing the higher testing accuracy. The full list of 3D patches from the left atrium was then classified using the optimal 3D CNN model, and the receiver operating curves (ROCs) were produced. The final average area under the curve (AUC) from the ROC curves was 0.840 ± 0.065 and the average accuracy was 78.9% ± 5.9%. This demonstrates that the CNN-based method is capable of distinguishing healthy hearts from those with previous cardiovascular disease.
关键词： Deep learning,Heart disease,Classification,Cardiovascular disease (CVD),Convolutional neural networks,Computer-aided diagnosis,3D Computed tomography
更新于2025-09-19 17:15:36