Lightweight facial landmark detection network based on improved MobileViT

被引:5
|
作者
Song, Limei [1 ]
Hong, Chuanfei [1 ]
Gao, Tian [1 ]
Yu, Jiali [1 ]
机构
[1] Tiangong Univ, Tianjin Key Lab Intelligent Control Elect Equipmen, Tianjin, Peoples R China
关键词
Facial landmark; MobileViT; Lightweight; Transformer;
D O I
10.1007/s11760-023-02975-4
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
The long-distance link between facial landmarks cannot be modeled by the current CNN-based facial landmark detection networks, and these networks typically have many parameters that consume substantial computational resources. This paper proposes a multi-scale lightweight facial landmark detection network with CNN and Transformer multi-branch parallelism. Based on MobileViT, the network incorporates MobileOne Block and simplified Ghost BottleNeck lightweight network structure. Compared to MobileViT on the WFLW dataset, the number of network parameters is reduced by 49.18%, the failure rate is reduced by 3.20%, the detection speed is improved by 41.73%, the FLOPS is reduced by 64.83%, and the NME is improved by 0.45% and 1.31% on the test and pose subsets, respectively. The data proves that the global information extraction of facial landmarks is more accurate after adding the Transformer structure. This paper also compares with other networks, and the result shows that improved MobileViT achieves more accurate detection with fewer model parameters.
引用
收藏
页码:3123 / 3131
页数:9
相关论文
共 50 条
  • [1] Lightweight facial landmark detection network based on improved MobileViT
    Limei Song
    Chuanfei Hong
    Tian Gao
    Jiali Yu
    Signal, Image and Video Processing, 2024, 18 : 3123 - 3131
  • [2] Recurrent neural network for facial landmark detection
    Chen, Yu
    Yang, Jian
    Qian, Jianjun
    NEUROCOMPUTING, 2017, 219 : 26 - 38
  • [3] A Marine Benthic Detection Algoritm Based On Improved Lightweight Network
    Le, Bai
    Ying, Hu
    Kun, Chang Hai
    2022 34TH CHINESE CONTROL AND DECISION CONFERENCE, CCDC, 2022, : 5795 - 5800
  • [4] Detection of early bruises in apples using hyperspectral imaging and an improved MobileViT network
    Yang, Mianqing
    Chen, Guoliang
    Lv, Feng
    Ma, Yunyun
    Wang, Yiyun
    Zhao, Qingdian
    Liu, Dayang
    JOURNAL OF FOOD SCIENCE, 2024,
  • [5] A lightweight Hardware Trojan detection approach in the waveform diagram based on MobileViT and attention mechanism
    Chen, Shouhong
    Qin, Guanxiang
    Lu, Ying
    Wang, Tao
    Hou, Xingna
    JOURNAL OF SUPERCOMPUTING, 2025, 81 (04):
  • [6] Light weight recognition of weeds in the field based on improved MobileViT network
    Yang S.
    Zhang H.
    Xing L.
    Du Y.
    Nongye Gongcheng Xuebao/Transactions of the Chinese Society of Agricultural Engineering, 2023, 39 (09): : 152 - 160
  • [7] Improved YOLOv5 Based on the Mobilevit Backbone for the Detection of Steel Surface Defects Improved YOLOv5 based on the mobilevit backbone and BiFPN
    Qiu, Kun
    Wang, Changkun
    PROCEEDINGS OF 2024 3RD INTERNATIONAL CONFERENCE ON CYBER SECURITY, ARTIFICIAL INTELLIGENCE AND DIGITAL ECONOMY, CSAIDE 2024, 2024, : 305 - 309
  • [8] An Improved Lightweight Parameters Network for Strawberry Flowers Detection
    Zhou, Bao
    Lin, Xueying
    Zhou, Jie
    Wang, Yujin
    Hu, Fangchao
    IEEE ACCESS, 2023, 11 : 63761 - 63772
  • [9] MVT-CEAM: a lightweight MobileViT with channel expansion and attention mechanism for facial expression recognition
    Wang, Kunxia
    Yu, Wancheng
    Yamauchi, Takashi
    SIGNAL IMAGE AND VIDEO PROCESSING, 2024, 18 (10) : 6853 - 6865
  • [10] A Lightweight Network Based on Improved YOLOv5s for Insulator Defect Detection
    Liu, Cong
    Yi, Wentao
    Liu, Min
    Wang, Yifeng
    Hu, Sheng
    Wu, Minghu
    ELECTRONICS, 2023, 12 (20)