Adaptive Long-neck Network with Atrous-Residual Structure for Instance Segmentation | |
Wenjie Geng; Zhiqiang Cao; Peiyu Guan; Guangli Ren; Junzhi Yu; Fengshui Jing | |
刊名 | IEEE Sensors Journal |
2023 | |
卷号 | 7期号:23页码:7786-7797 |
英文摘要 | Instance segmentation is an important yet challenging task in computer vision field. Existing mainstream single-stage solution with parameterized mask representation has designed the neck models to fuse features of different layers; however, the performance of instance segmentation is still restricted to the layer-bylayer transmission scheme. In this paper, an instance segmentation framework with an adaptive long-neck network and atrous-residual structure is proposed. The long-neck network is composed of two bi-directional fusion units, which are cascaded to facilitate the information communication among features of different layers in top-down and bottom-up pathways. Specially, a new cross-layer transmission scheme is introduced in top-down pathway to achieve hybrid dense fusion of multi-scale features and weights of different features are learned adaptively according to their respective contributions to promote the network convergence. Meanwhile, a bottom-up pathway further complements the features with more location clues. In this way, high-level semantic information and low-level location information are tightly integrated. Furthermore, an atrous-residual structure is added to the mask prototype branch of instance prediction to capture more contextual information. This contributes to the generation of high-quality masks. The experiment results indicate that the proposed method achieves effective segmentation and the outputted masks match the contours of objects. |
语种 | 英语 |
内容类型 | 期刊论文 |
源URL | [http://ir.ia.ac.cn/handle/173211/52253] |
专题 | 智能机器人系统研究 |
通讯作者 | Zhiqiang Cao |
推荐引用方式 GB/T 7714 | Wenjie Geng,Zhiqiang Cao,Peiyu Guan,et al. Adaptive Long-neck Network with Atrous-Residual Structure for Instance Segmentation[J]. IEEE Sensors Journal,2023,7(23):7786-7797. |
APA | Wenjie Geng,Zhiqiang Cao,Peiyu Guan,Guangli Ren,Junzhi Yu,&Fengshui Jing.(2023).Adaptive Long-neck Network with Atrous-Residual Structure for Instance Segmentation.IEEE Sensors Journal,7(23),7786-7797. |
MLA | Wenjie Geng,et al."Adaptive Long-neck Network with Atrous-Residual Structure for Instance Segmentation".IEEE Sensors Journal 7.23(2023):7786-7797. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论