DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion
In view of the lack of feature complementarity between the feature layers of Single Shot MultiBox Detector (SSD) and the weak detection ability of SSD for small objects, we propose an improved SSD object detection algorithm based on Dense Convolutional Network (DenseNet) and feature fusion, which is...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8978787/ |
| _version_ | 1828890570378969088 |
|---|---|
| author | Sheping Zhai Dingrong Shang Shuhuan Wang Susu Dong |
| author_facet | Sheping Zhai Dingrong Shang Shuhuan Wang Susu Dong |
| author_sort | Sheping Zhai |
| collection | DOAJ |
| description | In view of the lack of feature complementarity between the feature layers of Single Shot MultiBox Detector (SSD) and the weak detection ability of SSD for small objects, we propose an improved SSD object detection algorithm based on Dense Convolutional Network (DenseNet) and feature fusion, which is called DF-SSD. On the basis of SSD, we design the feature extraction network DenseNet-S-32-1 with reference to the dense connection of DenseNet, and replace the original backbone network VGG-16 of SSD with DenseNet-S-32-1 to enhance the feature extraction ability of the model. In the part of multi-scale detection, a fusion mechanism of multi-scale feature layers is introduced to organically combine low-level visual features and high-level semantic features in the network structure. Finally, a residual block is established before the object prediction to further improve the model performance. We train the DF-SSD model from scratch. The experimental results show that our model DF-SSD with 300 × 300 input achieves 81.4% mAP, 79.0% mAP, and 29.5% mAP on PASCAL VOC 2007, VOC 2012, and MS COCO datasets, respectively. Compared with SSD, the detection accuracy of DF-SSD on VOC 2007 is improved by 3.1% mAP. DF-SSD requires only 1/2 parameters to SSD and 1/9 parameters to Faster RCNN. We inject more semantic information into DF-SSD, which makes it have advanced detection effect on small objects and objects with specific relationships. |
| first_indexed | 2024-12-13T13:01:22Z |
| format | Article |
| id | doaj.art-d048cefa907a494597f90693429bec5c |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-13T13:01:22Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-d048cefa907a494597f90693429bec5c2022-12-21T23:44:59ZengIEEEIEEE Access2169-35362020-01-018243442435710.1109/ACCESS.2020.29710268978787DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature FusionSheping Zhai0https://orcid.org/0000-0001-8937-433XDingrong Shang1https://orcid.org/0000-0002-4740-8068Shuhuan Wang2https://orcid.org/0000-0002-1699-3612Susu Dong3School of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an, ChinaSchool of Computer Science and Technology, Xi’an University of Posts and Telecommunications, Xi’an, ChinaIn view of the lack of feature complementarity between the feature layers of Single Shot MultiBox Detector (SSD) and the weak detection ability of SSD for small objects, we propose an improved SSD object detection algorithm based on Dense Convolutional Network (DenseNet) and feature fusion, which is called DF-SSD. On the basis of SSD, we design the feature extraction network DenseNet-S-32-1 with reference to the dense connection of DenseNet, and replace the original backbone network VGG-16 of SSD with DenseNet-S-32-1 to enhance the feature extraction ability of the model. In the part of multi-scale detection, a fusion mechanism of multi-scale feature layers is introduced to organically combine low-level visual features and high-level semantic features in the network structure. Finally, a residual block is established before the object prediction to further improve the model performance. We train the DF-SSD model from scratch. The experimental results show that our model DF-SSD with 300 × 300 input achieves 81.4% mAP, 79.0% mAP, and 29.5% mAP on PASCAL VOC 2007, VOC 2012, and MS COCO datasets, respectively. Compared with SSD, the detection accuracy of DF-SSD on VOC 2007 is improved by 3.1% mAP. DF-SSD requires only 1/2 parameters to SSD and 1/9 parameters to Faster RCNN. We inject more semantic information into DF-SSD, which makes it have advanced detection effect on small objects and objects with specific relationships.https://ieeexplore.ieee.org/document/8978787/DenseNetfeature fusionmulti-scale object detectionSSD |
| spellingShingle | Sheping Zhai Dingrong Shang Shuhuan Wang Susu Dong DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion IEEE Access DenseNet feature fusion multi-scale object detection SSD |
| title | DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion |
| title_full | DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion |
| title_fullStr | DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion |
| title_full_unstemmed | DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion |
| title_short | DF-SSD: An Improved SSD Object Detection Algorithm Based on DenseNet and Feature Fusion |
| title_sort | df ssd an improved ssd object detection algorithm based on densenet and feature fusion |
| topic | DenseNet feature fusion multi-scale object detection SSD |
| url | https://ieeexplore.ieee.org/document/8978787/ |
| work_keys_str_mv | AT shepingzhai dfssdanimprovedssdobjectdetectionalgorithmbasedondensenetandfeaturefusion AT dingrongshang dfssdanimprovedssdobjectdetectionalgorithmbasedondensenetandfeaturefusion AT shuhuanwang dfssdanimprovedssdobjectdetectionalgorithmbasedondensenetandfeaturefusion AT susudong dfssdanimprovedssdobjectdetectionalgorithmbasedondensenetandfeaturefusion |