A Study on Small-Scale Ship Detection Based on Attention Mechanism

Zheng, Jianli; Liu, Yang

doi:10.1109/access.2022.3193669

Cited by 4 publications

(5 citation statements)

References 40 publications

(43 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in Figure 8, after image enhancement, it is difficult for the model to detect the ship objects in environments such as darkness, noisy, and cutout, resulting in a decrease in the model R of 4.83%, and the detection accuracy To verify that our method has a good detection effect, we compare it with previous SOTA approaches in the SeaShips dataset. The experimental models for comparison are small ship detection method [42], ShipYOLO [32], and enhanced YOLOv3-tiny [31]. The results are as shown.…”

Section: Test Of Model Interferencementioning

confidence: 89%

An Improved Ship Classification Method Based on YOLOv7 Model with Attention Mechanism

Cen

Feng

Liu

et al. 2023

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

Deep learning (DL) is widely used in ship detection, but there are still some problems in the effective classification, such as inaccurate object feature extraction and inconspicuous feature information in deep layers. To address these problems, we propose a YOLOv7-residual convolutional block attention module (YOLOv7-RCBAM) by combining the convolutional attention mechanism and residual connections to the YOLOv7. First, to accelerate the training speed, the parameters in the backbone network of the pretrained model are frozen by using transfer learning, and the model is fine-tuned for training. Second, to enhance the information relevance of channel dimensional features, an attention mechanism with residual connectivity is adopted. Finally, a feature fusion attention mechanism is introduced to improve the effective feature extraction. The effectiveness of the proposed method is fully validated on the SeaShips dataset. The results show that the YOLOv7-RCBAM model achieves better performance with a 97.59% in mAP and effectively extracts object feature in deep layers. Meanwhile, the YOLOv7-RCBAM model can accurately locate ship in complex environments with darkness and noise with the mAP reaching 96.13% to achieve effective ship classification detection.

show abstract

Section: Test Of Model Interferencementioning

confidence: 89%

An Improved Ship Classification Method Based on YOLOv7 Model with Attention Mechanism

Cen

Feng

Liu

et al. 2023

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

show abstract

“…Light_SDNet is based on YOLO5, and its result is also promising. Liu_2020 [39] and Liu_2022 [36] are based on older versions in the YOLO family; hence, the performance is smaller than the SoTA framework as Light_SDNet and our method. Second, the parameterization adds some uncertainty to the training process.…”

Section: Compare With Sotamentioning

confidence: 97%

“…Several modifications of well-known object detection methods have been introduced to improve the performance of ship detectors. Liu_2022 [36] based on the SSD [8] framework and VGG backbone to detect a ship on small scales. The author [36] used a local attention network to fuse cross-features; also, a merge module combines features from different scales to improve detection results.…”

Section: Ship Detectionmentioning

confidence: 99%

“…Liu_2022 [36] based on the SSD [8] framework and VGG backbone to detect a ship on small scales. The author [36] used a local attention network to fuse cross-features; also, a merge module combines features from different scales to improve detection results. The YOLO family is also used by many works to enhance the detection of ship datasets.…”

Section: Ship Detectionmentioning

confidence: 99%

“…Given 70,000 published images from SeaShip [19], Zhang_2022 [47], and Zhang_2021 [42] used 90% of data for training and validating; the other 10% of data are the testing dataset. Liu_2020 [39], Liu_2022 [36], Han_2021 [44], and Light_SDNet [21] use 80% of data for training and validating dataset, the other 20% is the testing dataset. To compare with these works, we use D Train 5 shows that our method is better than other methods in terms of average precision.…”

Section: Compare With Sotamentioning

confidence: 99%

See 2 more Smart Citations

Image-Based Ship Detection Using Deep Variational Information Bottleneck

Ngo,

Vo,

Nguyen

et al. 2023

Sensors

View full text Add to dashboard Cite

Image-based ship detection is a critical function in maritime security. However, lacking high-quality training datasets makes it challenging to train a robust supervision deep learning model. Conventional methods use data augmentation to increase training samples. This approach is not robust because the data augmentation may not present a complex background or occlusion well. This paper proposes to use an information bottleneck and a reparameterization trick to address the challenge. The information bottleneck learns features that focus only on the object and eliminate all backgrounds. It helps to avoid background variance. In addition, the reparameterization introduces uncertainty during the training phase. It helps to learn more robust detectors. Comprehensive experiments show that the proposed method outperforms conventional methods on Seaship datasets, especially when the number of training samples is small. In addition, this paper discusses how to integrate the information bottleneck and the reparameterization into well-known object detection frameworks efficiently.

show abstract

Orientation-aware ship detection via a rotation feature decoupling supported deep learning approach

Chen

Han

et al. 2023

Engineering Applications of Artificial Intelligence

View full text Add to dashboard Cite

A Study on Small-Scale Ship Detection Based on Attention Mechanism

Abstract: Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.

Cited by 4 publications

References 40 publications

An Improved Ship Classification Method Based on YOLOv7 Model with Attention Mechanism

An Improved Ship Classification Method Based on YOLOv7 Model with Attention Mechanism

Image-Based Ship Detection Using Deep Variational Information Bottleneck

Orientation-aware ship detection via a rotation feature decoupling supported deep learning approach

Contact Info

Product

Resources

About