In order to intelligently and efficiently identify the standardization of marketing on-site operations, an attention mechanism based method for identifying standardized small targets in marketing on-site operation images is proposed. In this method, a standardized small target recognition method based on attention mechanism for marketing on-site operation images is proposed to address the characteristics of image occlusion, diverse targets, and complex backgrounds. In the target detection network, attention mechanism has been added, detection head module has been added, and loss function has been optimized to improve the accuracy of model recognition. The proposed method achieves a 5.56% improvement in AP and a 5.35% improvement in mAP. We have collected and produced a large number of proprietary industry datasets for marketing on-site scenarios. The experimental results show that this method has high accuracy and robustness in identifying the standardization of marketing on-site operation images. At the same time, this method can provide effective technical means for the management and supervision of marketing on-site operations, and has high practical and promotional value.
JiangMWangYYuF, et al.UAV-FDN: forest-fire detection network for unmanned aerial vehicle perspective. J Intell Fuzzy Syst2023; 45(4): 5821–5836.
2.
LiuLZhouXHeM, et al.A real-time semantic segmentation network for robotic arm graspComputer graphics international conference. Springer Nature Switzerland, 2024, pp. 179–191.
3.
KulkarniMBRajagopalSPrieto-SimónB, et al.Recent advances in smart wearable sensors for continuous human health monitoring. Talanta2024; 272: 125817.
4.
JhaRMishraPKumarS. Advancements in optical fiber-based wearable sensors for smart health monitoring. Biosens Bioelectron2024; 254: 116232.
5.
YuFLiuJYuH, et al.Multimodal wearable system with dual-frequency enhancement network for risk recognition. IEEE Internet Things J2025; 12(11): 17634–17648.
6.
HuangYLiuHChenQ, et al.Transmission line insulator fault detection method based on USRNet and improved YOLOv5x. High Volt Eng2022; 48(9): 3437–3446.
7.
JinYXiaoZYuanJ, et al.MFENet: multi-Scale and local frequency enhancement network for skin lesion Classification Computer graphics international conference. Springer Nature Switzerland, 2024, pp. 192–203.
8.
DuCWangJYuF, et al.Latent diffusion-enhanced virtual Try-On via optimized pseudo-label generation. Proc AAAI Conf Artif Intell2025; 39(3): 2780–2788.
9.
TangWJiangXLiuH, et al.Lithium-ion battery fire early warning based on gas-liquid escape image recognition. High Voltage Technology2022; 48(8): 3295–3304.
10.
CaoJRenLDengA, et al.MHC-Segnet: mamba–hadamard collaboration segmentation network for multimodal MRI brain tumor. Vis Comput2025; 1–12.
11.
KaputaDSLandyBP. YOLBO: you only look back Once–a low latency object tracker based on YOLO and optical flow. IEEE Access2021; 9: 82497–82507.
12.
YuFXiaoZLiuL, et al.BiaCanDet: bioelectrical impedance analysis for breast cancer detection with space–time attention neural network. Expert Syst Appl2025; 269: 126223.
13.
HanZLiuJ. Image tamper detection algorithm based on improved YOLOX-X. Yangtze River Information and Communication2023; 36(01): pp. 84–86.
14.
LiSWangXZhangB, et al.Research on side scan sonar image sunken ship detection method based on improved YOLOX. Hydrographic Surveying and Charting2022; 42(05): 32–36.
15.
BuiVLeNTNguyenVH, et al.Multi-behavior with bottleneck features LSTM for load forecasting in building energy management system. Electronics2021; 10(9): 1026.
16.
ZhangYXieFHuangL, et al.A lightweight one-stage defect detection network for small object based on dual attention mechanism and PAFPN. Front Physiol2021; 9: 708097.
17.
WangZZXieKZhangXY, et al.Small-object detection based on yolo and dense block via image super-resolution. IEEE Access2021; 9: 56416–56429.
18.
JoshyAARajanR. Dysarthria severity assessment using squeeze-and-excitation networks. Biomed Signal Process Control2023; 82: 104606.
19.
DingJLiWPeiL, et al.Sw-YoloX: an anchor-free detector based transformer for sea surface object detection. Expert Syst Appl2023; 217: 119560.
20.
SunD. Analysis and implementation of image recognition technology based on BP neural network genetic algorithm. Information Technology and Informatization2022; 05: 190–193.
21.
DongYWanWHuW, et al.3D multi-target detection based on deformable convolution and data augmentation. Industrial Control Computer2023; 36(03): 22–24.
22.
ZhouWJinJLeiJ, et al.CEGFNet: common extraction and gate fusion network for scene parsing of remote sensing images. IEEE Trans Geosci Rem Sens2021; 60: 1–10.
23.
ZhongqiangWXueqinL. Microgrid fault diagnosis based on deep transfer learning and LSTM networ k. Acta Metrol Sin2023; 44(04): 582–590.
24.
BianJMaoZJiangB, et al.A fault diagnosis method for key equipment in industrial production based on knowledge graph and multi - task learning. Sci Sin Inform2023; 53(04): 699–714.
25.
TianYZhangYZhangH. Recent advances in stochastic gradient descent in deep learning. Mathematics2023; 11(3): 682.
26.
ArouriYSayyafzadehM. An adaptive moment estimation framework for well placement optimization. Comput Geosci2022; 26(4): 957–973.
27.
YuFYuCTianZ, et al.Intelligent wearable system with motion and emotion recognition based on digital twin technology. IEEE Internet Things J2024; 11(15): 26314–26328.
28.
XieJPengYWangM. The squeeze & excitation normalization based nnu-net for segmenting head & neck tumors. Chin J Electron2024; 33(3): 766–775.
29.
LiHWangYYiP, et al.Research on safety helmet recognition in complex work scenarios based on deep learning. Journal of Safety Science and Technology2021; 17(01): 175–181.
30.
BhattaraiMMartinez-RamonM. A deep learning framework for detection of targets in thermal images to improve firefighting. IEEE Access2020; 8: 88308–88321.
31.
JovicicEPrimoracDCupicM, et al.Publicly available datasets for predictive maintenance in the energy sector: a review. IEEE Access2023; 11: 73505–73520.
32.
DengYCaoKKHuW, et al.Harmonized and open energy dataset for modeling a highly renewable Brazilian power system. Sci Data2023; 10(1): 103.
33.
WangSZhangCSuD, et al.High-precision binary object detector based on a bsf-xnor convolutional layer. IEEE Access2021; 9: 106169–106180.
34.
IrieKNishikawaK. Detection method from 4k images using ssd300 without retraining[C]//2022 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC). IEEE, 2022, pp. 877–883.
35.
EggertCBrehmSWinschelA, et al.A closer look: small object detection in faster R-CNN. In: 2017 IEEE international conference on multimedia and expo (ICME). IEEE, 2017, pp. 421–426.
36.
HnewaMRadhaH. Multiscale domain adaptive yolo for cross-domain object detection. In: 2021 IEEE international conference on image processing (ICIP). IEEE, 2021, pp. 3323–3327.
