A CNN-Based Spatial Feature Fusion Algorithm for Hyperspectral Imagery Classification

Guo, Alan J. X.; Zhu, Fei

doi:10.1109/tgrs.2019.2911993

Cited by 38 publications

(19 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Deep learning is successful in diverse applications, such as image retrieval, speech recognition, and biomedical signal classification [24]- [28]. It is noteworthy that convolutional neural networks (CNNs) have gained substantial interest due to their strong spatial feature extraction ability [29]- [32], which might be promising algorithms for P300 detection [33], [34]. Cecotti first used CNN to realize the classification of P300 and achieved a high recognition rate (90%) with 10 repetitions [35].…”

Section: Introductionmentioning

confidence: 99%

Performance Enhancement of P300 Detection by Multiscale-CNN

Wang

Pei

et al. 2021

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

P300-based spelling system is one of the most popular brain-computer interface applications. Its success largely depends on performance, including the information transmission rate (ITR) and detection rate (i.e., accuracy). To achieve good performance, we proposed a multi-scale convolutional neural network (MS-CNN) model, which consists of seven layers. First, an upfront dataset was used to train the MS-CNN, aiming to obtain a subject-unspecific model (universal model) for P300 detection. Second, this universal model was adapted by a portion of data derived from a subject to update the model to obtain a subject-specific model by incorporating a transfer learning technique. We applied the proposed model in the BCI Controlled Robot Contest at the 2019 World Robot Conference, and our performance was the best among the teams in the contest. In the contest, ten healthy young subjects were randomly assigned by the contest committee to assess the model. Our model achieved the best P300 detection performance (higher accuracy with less repetition time). The ITR for the subject-unspecific case was 13.49 bits/min while the ITR for the subject-specific case was 12.13 bits/min when the repetitions were fewer than six. It is believed that our method may pave a promising path for taking a further step toward efficient implementation of P300-based spelling system.

show abstract

Section: Introductionmentioning

confidence: 99%

Performance Enhancement of P300 Detection by Multiscale-CNN

Wang

Pei

et al. 2021

IEEE Trans. Instrum. Meas.

View full text Add to dashboard Cite

show abstract

“…For instance, support vector machine (SVM) was applied to HSI data for strawberry ripeness evaluation achieving classification accuracy over 85%. Convolutional neural network (CNN), being the current state-of-the-art in deep learning [ 15 ], first achieved success in the field of image recognition and has become an extremely popular tool for remotely sensed HSI data classification [ 10 ]. More importantly, CNN models show flexibility to deal with HSI data by introducing a one-dimensional CNN for processing spectral inputs [ 16 ], two-dimensional CNN for single or multiple wavelength images [ 17 ], and three-dimensional CNN (3D-CNN) for an intelligent combination of spectral and spatial image data [ 18 , 19 ].…”

Section: Introductionmentioning

confidence: 99%

Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products

Zhu

Gowen

Feng³

et al. 2020

Sensors

View full text Add to dashboard Cite

Hyperspectral imaging (HSI) emerges as a non-destructive and rapid analytical tool for assessing food quality, safety, and authenticity. This work aims to investigate the potential of combining the spectral and spatial features of HSI data with the aid of deep learning approach for the pixel-wise classification of food products. We applied two strategies for extracting spatial-spectral features: (1) directly applying three-dimensional convolution neural network (3-D CNN) model; (2) first performing principal component analysis (PCA) and then developing 2-D CNN model from the first few PCs. These two methods were compared in terms of efficiency and accuracy, exemplified through two case studies, i.e., classification of four sweet products and differentiation between white stripe (“myocommata”) and red muscle (“myotome”) pixels on salmon fillets. Results showed that combining spectral-spatial features significantly enhanced the overall accuracy for sweet dataset, compared to partial least square discriminant analysis (PLSDA) and support vector machine (SVM). Results also demonstrated that spectral pre-processing techniques prior to CNN model development can enhance the classification performance. This work will open the door for more research in the area of practical applications in food industry.

show abstract

“…Existed CNN based detection methods mostly calculate false alarm based on the result itself, e.g., SAR image detection [36], while most works do not consider how to achieve better detection performance with controllable false alarms. Therefore, further research on the CNN-based detection method is needed to meet the actual radar false alarm requirement [37][38][39].…”

Section: Introductionmentioning

confidence: 99%

False-Alarm-Controllable Radar Detection for Marine Target Based on Multi Features Fusion via CNNs

Chen

Huang

et al. 2021

IEEE Sensors J.

View full text Add to dashboard Cite

Due to the influence of the complex marine environment, the marine target detection based on statistical theory is difficult to achieve high-performance. Moreover, due to various targets' motion characteristics, only using a single feature for detection is unreliable. In this paper, from the perspective of feature extraction and classification, marine target and sea clutter are classified by deep learning methods. To achieve the required false alarm rate, the dual-channel convolutional neural networks (DCCNN) and false-alarm-controllable classifier (FACC)-based marine target detection method is proposed. Firstly, the measured sea clutter and the target signal are preprocessed to obtain the time-Doppler spectrum and amplitude information. The Marine-DCCNN (MDCCNN) is then constructed for features extraction and fusion, and the feature vectors of the signals are obtained. The performance of different feature extraction models is tested and compared. Finally, the FACC is used as a detector to classify the feature vectors into two categories and the control of the false alarm rate is realized. The detection performances were verified by two popular public radar datasets, i.e., IPIX radar dataset (floating target) and CSIR dataset (maneuvering marine target). The results show that compared with single-channel CNN and histogram of oriented gradient support vector machine (Hog-SVM) classification, a combination of MDCCNN feature extraction model and softmax classifier can achieve higher performance and controllable false alarm rate. Moreover, HH polarization and mixed training datasets under different sea states can help improve detection performance. Index Terms-Radar target detection; marine target; feature extraction; dual-channel convolutional neural network (DCCNN); false alarm controllable classifier (FACC)

show abstract

A CNN-Based Spatial Feature Fusion Algorithm for Hyperspectral Imagery Classification

Cited by 38 publications

References 49 publications

Performance Enhancement of P300 Detection by Multiscale-CNN

Performance Enhancement of P300 Detection by Multiscale-CNN

Deep Spectral-Spatial Features of Near Infrared Hyperspectral Images for Pixel-Wise Classification of Food Products

False-Alarm-Controllable Radar Detection for Marine Target Based on Multi Features Fusion via CNNs

Contact Info

Product

Resources

About