Scene Classification Based on Multiscale Convolutional Neural Network

Liu, Yaqi; Guan, Qingxiao; Zhao, Xianfeng; Cao, Yun

doi:10.1109/tgrs.2018.2848473

Cited by 139 publications

(107 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Blackouts caused by large scale hazards such as hurricanes have been studied in e.g. [5]- [13]. In contrast, we focus on more localised phenomenons related to convective storms.…”

Section: Introductionmentioning

confidence: 99%

Short-Term Prediction of Electricity Outages Caused by Convective Storms

Tervo

Karjalainen

Jung

2019

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

Prediction of power outages caused by convective storms which are highly localised in space and time is of crucial importance to power grid operators. We propose a new machine learning approach to predict the damage caused by storms. This approach hinges identifying and tracking of storm cells using weather radar images on the application of machine learning techniques. Overall prediction process consists of identifying storm cells from CAPPI weather radar images by contouring them with a solid 35 dBZ threshold, predicting a track of storm cells and classifying them based on their damage potential to power grid operators. Tracked storm cells are then classified by combining data obtained from weather radar, ground weather observations and lightning detectors. We compare random forest classifiers and deep neural networks as alternative methods to classify storm cells. The main challenge is that the training data are heavily imbalanced as extreme weather events are rare.

show abstract

“…Blackouts caused by large scale hazards such as hurricanes have been studied in e.g. [5]- [13]. In contrast, we focus on more localised phenomenons related to convective storms.…”

Section: Introductionmentioning

confidence: 99%

Short-Term Prediction of Electricity Outages Caused by Convective Storms

Tervo

Karjalainen

Jung

2019

IEEE Trans. Geosci. Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Conventional RS image scene classification/retrieval approaches rely on several low-level features such as shape, texture, spectral information [4,5]. Recent advances in deep learning show that Convolutional Neural Networks (CNNs) lead to very high scene classification performance due to their high capability to model high-level semantic content of RS images [6,7]. In recent years, CNN architectures such as GoogLeNet [8], CaffeNet [9] have shown to achieve state-of-the-art classification performance for RS images [10].…”

Section: Introductionmentioning

confidence: 99%

“…In [11], three strategies i.e., using pre-training, fine-tuning and using CNN as feature extractor were explored to RS scene classification problems. In [12], a scale invariant CNN was introduced to avoid discriminative information loss during scene classification that are usually obtained while using fixed-scale images. In [13], a two-tunnel CNN approach was introduced to perform scene classification for multi-source RS images.…”

Section: Introductionmentioning

confidence: 99%

“…To reduce the storage size of the RS image archives, RS images are usually stored in compressed format. However, most of the existing RS image scene classification approaches in deep learning requires full decompression of images since CNNs do not directly operate on the compressed streams of RS images [10][11][12][13][14]. Thus, decoding of the images is required, which is computationally demanding and time-consuming for operational RS applications on large-scale RS image archives.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Approximating JPEG 2000 wavelet representation through deep neural networks for remote sensing image scene classification

Byju

Sümbül

Demir

et al. 2019

Image and Signal Processing for Remote Sensing XXV

View full text Add to dashboard Cite

This paper presents a novel approach based on the direct use of deep neural networks to approximate wavelet subbands for remote sensing (RS) image scene classification in the JPEG 2000 compressed domain. The proposed approach consists of two main steps. The first step aims to approximate the finer level wavelet sub-bands. To this end, we introduce a novel Deep Neural Network approach that utilizes the coarser level binary decoded wavelet sub-bands to approximate the finer level wavelet sub-bands (the image itself) through a series of deconvolutional layers. The second step aims to describe the high-level semantic content of the approximated wavelet subbands and to perform scene classification based on the learnt descriptors. This is achieved by: i) a series of convolutional layers for the extraction of descriptors which models the approximated sub-bands; and ii) fully connected layers for the RS image scene classification. Then, we introduce a loss function that allows to learn the parameters of both steps in an end-to-end trainable and unified neural network. The proposed approach requires only the coarser level wavelet sub-bands as input and thus minimizes the amount of decompression applied to the compressed RS images. Experimental results show the effectiveness of the proposed approach in terms of classification accuracy and reduced computational time when compared to the conventional use of Convolutional Neural Networks within the JPEG 2000 compressed domain.

show abstract

“…The literature [45] used an improved Vector of Locally Aggregated Descriptors (VLAD) algorithm to encode the deep regional features for generating local feature representation of remote sensing images. To obtain robust features to objects' scale, the literature [46,47] first fed the images with different scales produced by randomly stretching an image into ConvNets, then extracted the fully connected layer features. The discriminant correlation analysis algorithm was employed in [48] to fuse the last two fully connected layer features of pre-trained VggNet.…”

mentioning

confidence: 99%

A Multi-Scale Approach for Remote Sensing Scene Classification Based on Feature Maps Selection and Region Representation

Zhang

Shi

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Scene classification is one of the bases for automatic remote sensing image interpretation. Recently, deep convolutional neural networks have presented promising performance in high-resolution remote sensing scene classification research. In general, most researchers directly use raw deep features extracted from the convolutional networks to classify scenes. However, this strategy only considers single scale features, which cannot describe both the local and global features of images. In fact, the dissimilarity of scene targets in the same category may result in convolutional features being unable to classify them into the same category. Besides, the similarity of the global features in different categories may also lead to failure of fully connected layer features to distinguish them. To address these issues, we propose a scene classification method based on multi-scale deep feature representation (MDFR), which mainly includes two contributions: (1) region-based features selection and representation; and (2) multi-scale features fusion. Initially, the proposed method filters the multi-scale deep features extracted from pre-trained convolutional networks. Subsequently, these features are fused via two efficient fusion methods. Our method utilizes the complementarity between local features and global features by effectively exploiting the features of different scales and discarding the redundant information in features. Experimental results on three benchmark high-resolution remote sensing image datasets indicate that the proposed method is comparable to some state-of-the-art algorithms.

show abstract

Scene Classification Based on Multiscale Convolutional Neural Network

Cited by 139 publications

References 55 publications

Short-Term Prediction of Electricity Outages Caused by Convective Storms

Short-Term Prediction of Electricity Outages Caused by Convective Storms

Approximating JPEG 2000 wavelet representation through deep neural networks for remote sensing image scene classification

A Multi-Scale Approach for Remote Sensing Scene Classification Based on Feature Maps Selection and Region Representation

Contact Info

Product

Resources

About