Scanning Ion Conductance Microscopy (SICM) enables non-destructive imaging of living cells, which makes it highly valuable in life sciences, medicine, pharmacology, and many other fields. However, because of the uncer...
详细信息
Scanning Ion Conductance Microscopy (SICM) enables non-destructive imaging of living cells, which makes it highly valuable in life sciences, medicine, pharmacology, and many other fields. However, because of the uncertainty retrace height of SICM hopping mode, the time resolution of SICM is relatively low, which makes the device fail to meet the demands of dynamic scanning. To address above issues, we propose a fast-scanning method for SICM based on an autoencoder network. Firstly, we cut under-sampled images into small image lists. Secondly, we feed them into a self-constructed primitive-autoencoder super-resolution network to compute high-resolution images. Finally, the inferred scanning path is determined using the computed images to reconstruct the real high-resolution scanning path. The results demonstrate that the proposed network can reconstruct higher-resolution images in various super-resolution tasks of low-resolution scanned images. Compared to existing traditional interpolation methods, the average peak signal-to-noise ratio improvement is greater than 7.5823 dB, and the average structural similarity index improvement is greater than 0.2372. At the same time, using the proposed method for high-resolution image scanning leads to a 156.25% speed improvement compared to traditional methods. It opens up possibilities for achieving high-time resolution imaging of dynamic samples in SICM and further promotes the widespread application of SICM in the future.
This research explores the detection of flame front evolution in spark-ignition engines using an innovative neural network, the autoencoder. High-speed camera images from an optical access engine were analyzed under d...
详细信息
This research explores the detection of flame front evolution in spark-ignition engines using an innovative neural network, the autoencoder. High-speed camera images from an optical access engine were analyzed under different air excess coefficient lambda conditions to evaluate the autoencoder's performance. This study compared this new approach (AE) with an established method used by the same research group (BR) across multiple combustion cycles. Results revealed that the AE method outperformed the BR in accurately identifying flame pixels and significantly reducing overestimations outside the flame boundary. AE exhibited higher sensitivity levels, indicating its superior ability to identify pixels and minimize errors compared to the BR method. Additionally, AE's accuracy in representing combustion evolution was notably improved, offering a more detailed depiction of the process. AE's strength lies in its independence from specific threshold searches, a requirement in the BR method. By relying on learned representations within its latent space, AE eliminates laborious threshold exploration, ensuring reliability and reducing workload pressures. Comparative analyses consistently confirmed AE's superior performance in accurately reproducing and delineating combustion evolution compared to BR. This study highlights AE's potential as a promising technique for precise flame front detection in combustion processes. Its ability to autonomously extract features, minimize errors, and enhance overall accuracy signifies a significant step forward in analyzing flame fronts. AE's reliability, reduced need for manual intervention, and adaptability across various conditions suggest a promising future for improving combustion analysis techniques in spark-ignition engines with optical access.
This work addresses the challenge of the portability of autoencoder models for the lossy compression of different spatially independent and unknown hyperspectral satellite data. We propose an advanced 1D-Convolutional...
详细信息
ISBN:
(数字)9781665470698
ISBN:
(纸本)9781665470698
This work addresses the challenge of the portability of autoencoder models for the lossy compression of different spatially independent and unknown hyperspectral satellite data. We propose an advanced 1D-Convolutional autoencoder architecture for lossy hyperspectral data compression with high transferability to unknown spectral signatures. In the first experiment, the model is trained on a single PRISMA data set, and in the second experiment it is trained on five PRISMA data sets from all over the world. The abstraction ability of the two models is verified by processing six spatially independent hyperspectral PRISMA satellite data sets. The evaluation is based on the reconstruction accuracy using the SNR and SA metrics and compares it to other learning-based lossy compression techniques. We demonstrate the high transferability and generalization of our 1D-Convolutional autoencoder for a fixed compression ratio on each PRISMA satellite data set, which results in superior reconstruction accuracy compared to state-of-the-art methods.
This paper introduces a hybrid deep autoencoder with a random forest classifier model to enhance intrusion detection performance in a native SDN environment. A deep learning architecture combining a deep autoencoder w...
详细信息
ISBN:
(数字)9781538683477
ISBN:
(纸本)9781538683477
This paper introduces a hybrid deep autoencoder with a random forest classifier model to enhance intrusion detection performance in a native SDN environment. A deep learning architecture combining a deep autoencoder with random forest learning feature representation of traffic flows natively collected from the SDN environment. Publicly available packet Capture (PCAP) files of recorded traffic flows were used in the SDN network for flow feature extraction and real-time implementation. The results show very high and consistent performance metrics, with an average of 0.9 receiver-operating characteristics area under curve (ROC AUC) recorded. Furthermore, we compared the performance achieved using the original dataset with previous research to investigate the performance achieved using the same model developed. The flow-based intrusion model presented outperforms other publicly available methods, with a traffic anomaly detection rate of 98% accuracy and precision.
This paper considers reconfigurable intelligent surface (RIS)-assisted point-to-point multiple-input multiple-output (MIMO) communication systems, where a transmitter communicates with a receiver through an RIS. Based...
详细信息
ISBN:
(数字)9781665480536
ISBN:
(纸本)9781665480536
This paper considers reconfigurable intelligent surface (RIS)-assisted point-to-point multiple-input multiple-output (MIMO) communication systems, where a transmitter communicates with a receiver through an RIS. Based on the main target of reducing the bit error rate (BER) and therefore enhancing the communication reliability, we study different model-based and data-driven (autoencoder) approaches. In particular, we consider a model-based approach that optimizes both active and passive optimization variables. We further propose a novel end-to-end data-driven framework, which leverages the recent advances in machine learning. The neural networks presented for conventional signal processing modules are jointly trained with the channel effects to minimize the bit error detection. Numerical results demonstrate that the proposed data-driven approach can learn to encode the transmitted signal via different channel realizations dynamically. In addition, the data-driven approach not only offers a significant gain in the BER performance compared to the other state-of-the-art benchmarks but also guarantees the performance when perfect channel information is unavailable.
In this paper we present an accelerated algorithm for clustering source-dominated microphones in acoustic sensor networks. Predicated on privacy-preserving unsupervised clustered federated learning that groups microph...
详细信息
ISBN:
(数字)9781665468671
ISBN:
(纸本)9781665468671
In this paper we present an accelerated algorithm for clustering source-dominated microphones in acoustic sensor networks. Predicated on privacy-preserving unsupervised clustered federated learning that groups microphones by evaluating the similarity of model weight updates, we introduce a light-weight variational autoencoder and equip the algorithm with supplementary control criteria for faster convergence. We validate the quality, degree of acceleration and utility of our method using clustering-based and classification-based tasks. Compared to the previously employed deterministic autoencoder, we observe a significantly lower number of client-server communication rounds at the price of a minor reduction in clustering performance.
In adaptive learning, Cognitive diagnosis aims to discover students' knowledge state on different knowledge concepts and predict their future performance. Most previous methods consider more on students' own a...
详细信息
ISBN:
(纸本)9798350397444
In adaptive learning, Cognitive diagnosis aims to discover students' knowledge state on different knowledge concepts and predict their future performance. Most previous methods consider more on students' own answering history and rarely model the the impact brought by students with similar answering behaviors explicitly. This collaborative information among students is helpful for students who lack sufficient historical logs. In this paper, we propose a new cognitive diagnosis method called Dual autoencoder Enhanced Subgraph Pattern Mining(DASPM) for Cognitive Diagnosis, which incorporates collaborative information among students into the cognitive diagnosis process to obtain more accurate predictions. Specifically, we use a graph neural network to capture collaborative pattern on the student-exercise bipartite graph. In order to filter out the interference of irrelevant information, we design a subgraph extraction algorithm that separates local parts around the target student-exercise pair from global graph based on the correlation between exercises. In addition, we utilize a dual autoencoder module to encode students and exercises to enhance the initial representation of nodes in the subgraph. Extensive experiments on multiple datasets show the effectiveness of our proposed method.
In recent years, classification-based speech emotion recognition (SER) methods have achieved high overall performance. However, these methods tend to have lower performance for neutral speeches, which account for a la...
详细信息
In recent years, classification-based speech emotion recognition (SER) methods have achieved high overall performance. However, these methods tend to have lower performance for neutral speeches, which account for a large proportion in most practical situations. To solve the problem and improve the SER performance, we propose a neutral speech detector (NSD) based on the anomaly detection approach, which uses an autoencoder, the intermediate layer output of a pretrained SER classifier and only neutral data for training. The intermediate layer output of a pretrained SER classifier enables the reconstruction of both acoustic and text features, which are optimized for SER tasks. We then propose the combination of the SER classifier and the NSD used as a screening mechanism for correcting the class probability of the incorrectly recognized neutral speeches. Results of our experiment using the IEMOCAP dataset indicate that the NSD can reconstruct both the acoustic and textual features, achieving a satisfactory performance for use as a reliable screening method. Furthermore, we evaluated the performance of our proposed screening mechanism, and our experiments show significant improvement of 12.9% in the F-score of the neutral class to 80.3%, and 8.4% in the class-average weighted accuracy to 84.5% compared with state-of-the-art SER classifiers.
The variational autoencoder (VAE) has been used in a myriad of applications, e.g., dimensionality reduction and generative modeling. VAE uses a specific model for stochastic sampling in latent space. The normal distri...
详细信息
ISBN:
(纸本)9781665421607;9781665421591
The variational autoencoder (VAE) has been used in a myriad of applications, e.g., dimensionality reduction and generative modeling. VAE uses a specific model for stochastic sampling in latent space. The normal distribution is the most commonly used one because it allows a straightforward sampling, a reparameterization trick, and a differentiable expression of the Kullback-Leibler divergence. Although various other distributions such as Laplace were studied in literature, the effect of heterogeneous use of different distributions for posterior-prior pair is less known to date. In this paper, we investigate numerous possibilities of such a mismatched VAE, e.g., where the uniform distribution is used as a posterior belief at the encoder while the Cauchy distribution is used as a prior belief at the decoder. To design the mismatched VAE, the total number of potential combinations to explore grows rapidly with the number of latent nodes when allowing different distributions across latent nodes. We propose a novel framework called AutoVAE, which searches for better pairing set of posterior-prior beliefs in the context of automated machine learning for hyperparameter optimization. We demonstrate that the proposed irregular pairing offers a potential gain in the variational Renyi bound. In addition, we analyze a variety of likelihood beliefs and divergence order.
Out-of-town recommendation aims to provide Point-of-Interest (POI) recommendation when users leave their hometown and arrive in a new city. To infer the out-of-town preferences of cold-start users based on their homet...
详细信息
ISBN:
(纸本)9783031109898;9783031109881
Out-of-town recommendation aims to provide Point-of-Interest (POI) recommendation when users leave their hometown and arrive in a new city. To infer the out-of-town preferences of cold-start users based on their hometown preferences, some recent methods directly train a mapping function between users' hometown preferences and out-of-town preferences. Unfortunately, they depend on a large number of overlapping users who left check-in histories in both the home city and the out-of-town city to build the mapping relationships. Also, they don't fully explore the category hierarchy knowledge of POIs, which can help with robust POI representations. To this end, in this paper, we propose Adversarial Cycle-Consistent autoencoder for Category-Aware Out-of-Town Recommendation named ACCAC, which effectively learns the mapping function even in the case that the number of overlapping users is limited. Specifically, we first utilize denoising autoencoders to learn pre-trained POI embeddings augmented with category hierarchy knowledge. Then we introduce a cycle-consistent generative adversarial network to explore potential mapping relationships. Extensive experiments on real-world out-of-town recommendation datasets demonstrate the effectiveness of ACCAC.
暂无评论