Combing Single Shot Multibox Detector with transfer learning for ship detection using Chinese Gaofen-3 images

Wang, Yuanyuan; Wang, Chao; Zhang, Hong; Cheng, Zhang; Fu, Qiaoyan

doi:10.1109/piers-fall.2017.8293227

Cited by 27 publications

(12 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A good detection model is defined of high P d and low F a . The probabilities are formulated as [17] det…”

Section: Resultsmentioning

confidence: 99%

Combining of Transfer Learning with Faster-RCNN For Aedes Aegyti Larvae Detection

2019

IJRTE

View full text Add to dashboard Cite

The dengue epidemiology episode has become one of the global phenomena especially the rain forest countries including Malaysia. Environmental management, the used of chemical and biological environment are control strategies that has been proposed and practiced by World Health Organization. However, based on statistic al of dengue cases, there is still no concrete solution in curbing this problem especially at non-accessible places. This paper proposed a study on detection Aedes Aegypti larvae in water storage tank by combining transfer learning with Faster-RCNN. The purpose of the study is to acquire train and validation losses along with detection performance metrics. The experimental results disclose that the probability detection has scored 97.01% while false alarm has scored 5.97%. Those significant value has depicted that the trained model has high detection accuracies.

show abstract

“…A good detection model is defined of high P d and low F a . The probabilities are formulated as [17] det…”

Section: Resultsmentioning

confidence: 99%

Combining of Transfer Learning with Faster-RCNN For Aedes Aegyti Larvae Detection

2019

IJRTE

View full text Add to dashboard Cite

show abstract

“…This paper selects SSD512 [26] in the SSD series to make improvement. SSD512 provides deep convolutional neural network models of large, medium and small scales.…”

Section: Experimental Parameter Settingsmentioning

confidence: 99%

Multi-Object Detection in Traffic Scenes Based on Improved SSD

et al. 2018

View full text Add to dashboard Cite

In order to solve the problem that, in complex and wide traffic scenes, the accuracy and speed of multi-object detection can hardly be balanced by the existing object detection algorithms that are based on deep learning and big data, we improve the object detection framework SSD (Single Shot Multi-box Detector) and propose a new detection framework AP-SSD (Adaptive Perceive). We design a feature extraction convolution kernel library composed of multi-shape Gabor and color Gabor and then we train and screen the optimal feature extraction convolution kernel to replace the low-level convolution kernel of the original network to improve the detection accuracy. After that, we combine the single image detection framework with convolution long-term and short-term memory networks and by using the Bottle Neck-LSTM memory layer to refine and propagate the feature mapping between frames, we realize the temporal association of network frame-level information, reduce the calculation cost, succeed in tracking and identifying the targets affected by strong interference in video and reduce the missed alarm rate and false alarm rate by adding an adaptive threshold strategy. Moreover, we design a dynamic region amplification network framework to improve the detection and recognition accuracy of low-resolution small objects. Therefore, experiments on the improved AP-SSD show that this new algorithm can achieve better detection results when small objects, multiple objects, cluttered background and large-area occlusion are involved, thus ensuring this algorithm a good engineering application prospect.

show abstract

“…Also, an additional study [13] utilized the intermediate layers of a pre-trained network on a CIFAR-10 dataset as the feature extractor for the TerraSAR-X images segmentation. Wang et al [14] fine-tuned the VGG-16 model, which was trained under natural images, to detect ships in SAR images. Moreover, Marmanis et al [15] observed that initialization with the weights learned from optical images has little effect on the segmentation of SAR data because the distributions of optical images and SAR data are probably too different from each other for transferring, even in low layers.…”

Section: Introductionmentioning

confidence: 99%

Transfer Learning With CNNs for Segmentation of PALSAR-2 Power Decomposition Components

Vinayaraj

Sugimoto

Nakamura

et al. 2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Water/ice/land region segmentation is an important task for remote sensing, as it analyses the occurrence of water or ice on the earth's surface. Many previous deep learning researches effectively utilized multispectral satellite images for highly accurate water/ice/land region segmentation. However, the deep-learning-based segmentation of synthetic aperture radar (SAR) images still remains a challenging task due to the unavailability of enough labeled data. In order to overcome this issue, we designed a two-step deep-learning-based transfer learning model that needs a very limited number of labeled samples. The proposed approach consists of two models. The first model is a deep encoder-decoder 6SD to Landsat-8 multispectral translation model (DTF) that translates fully polarimetric PALSAR-2 6SD data to 6 new features. As for the second model (transfer learning), it utilizes the DTF features to fine-tune the model using the Landsat-8 multispectral pre-trained model for water/ice/land segmentation. Hereinafter, the proposed two-step model is referred to as DTF-TL. Also, a qualitative and quantitative analysis was carried out to evaluate the performance of the proposed model (DTF-TL) and compare it with various transfer learning methods. Overall, the DTF-TL model outperformed the other models with consistent and reliable water/ice/land segmentation results in terms of the recall (0.980), precision (0.981), F1-score (0.981), MIoU (0.962), and Accuracy (0.989).

show abstract

Combing Single Shot Multibox Detector with transfer learning for ship detection using Chinese Gaofen-3 images

Cited by 27 publications

References 7 publications

Combining of Transfer Learning with Faster-RCNN For Aedes Aegyti Larvae Detection

Combining of Transfer Learning with Faster-RCNN For Aedes Aegyti Larvae Detection

Multi-Object Detection in Traffic Scenes Based on Improved SSD

Transfer Learning With CNNs for Segmentation of PALSAR-2 Power Decomposition Components

Contact Info

Product

Resources

About