Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes
Nowadays, the commercial potential of live e-commerce is being continuously explored, and machine vision algorithms are gradually attracting the attention of marketers and researchers. During live streaming, the visuals can be effectively captured by algorithms, thereby providing additional data sup...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-09-01
|
Series: | Applied Sciences |
Subjects: | |
Online Access: | https://www.mdpi.com/2076-3417/13/18/10170 |
_version_ | 1797581458089443328 |
---|---|
author | Zongwei Li Kai Qiao Jianing Chen Zhenyu Li Yanhui Zhang |
author_facet | Zongwei Li Kai Qiao Jianing Chen Zhenyu Li Yanhui Zhang |
author_sort | Zongwei Li |
collection | DOAJ |
description | Nowadays, the commercial potential of live e-commerce is being continuously explored, and machine vision algorithms are gradually attracting the attention of marketers and researchers. During live streaming, the visuals can be effectively captured by algorithms, thereby providing additional data support. This paper aims to consider the diversity of live streaming devices and proposes an extremely lightweight and high-precision model to meet different requirements in live streaming scenarios. Building upon yolov5s, we incorporate the MobileNetV3 module and the CA attention mechanism to optimize the model. Furthermore, we construct a multi-object dataset specific to live streaming scenarios, including anchor facial expressions and commodities. A series of experiments have demonstrated that our model realized a 0.4% improvement in accuracy compared to the original model, while reducing its weight to 10.52%. |
first_indexed | 2024-03-10T23:04:59Z |
format | Article |
id | doaj.art-b04b16a70e7545f5844e408918c91a5d |
institution | Directory Open Access Journal |
issn | 2076-3417 |
language | English |
last_indexed | 2024-03-10T23:04:59Z |
publishDate | 2023-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Applied Sciences |
spelling | doaj.art-b04b16a70e7545f5844e408918c91a5d2023-11-19T09:23:58ZengMDPI AGApplied Sciences2076-34172023-09-0113181017010.3390/app131810170Improved Lightweight Multi-Target Recognition Model for Live Streaming ScenesZongwei Li0Kai Qiao1Jianing Chen2Zhenyu Li3Yanhui Zhang4School of Economics and Management, Shanghai Institute of Technology, Shanghai 200235, ChinaSchool of Economics and Management, Shanghai Institute of Technology, Shanghai 200235, ChinaSchool of Economics and Management, Shanghai Institute of Technology, Shanghai 200235, ChinaSchool of Cultural Heritage and Information Management, Shanghai University, Shanghai 200444, ChinaBusiness School, East China University of Science and Technology, Shanghai 200237, ChinaNowadays, the commercial potential of live e-commerce is being continuously explored, and machine vision algorithms are gradually attracting the attention of marketers and researchers. During live streaming, the visuals can be effectively captured by algorithms, thereby providing additional data support. This paper aims to consider the diversity of live streaming devices and proposes an extremely lightweight and high-precision model to meet different requirements in live streaming scenarios. Building upon yolov5s, we incorporate the MobileNetV3 module and the CA attention mechanism to optimize the model. Furthermore, we construct a multi-object dataset specific to live streaming scenarios, including anchor facial expressions and commodities. A series of experiments have demonstrated that our model realized a 0.4% improvement in accuracy compared to the original model, while reducing its weight to 10.52%.https://www.mdpi.com/2076-3417/13/18/10170model optimizationobject detectionattention mechanismlive streaming |
spellingShingle | Zongwei Li Kai Qiao Jianing Chen Zhenyu Li Yanhui Zhang Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes Applied Sciences model optimization object detection attention mechanism live streaming |
title | Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes |
title_full | Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes |
title_fullStr | Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes |
title_full_unstemmed | Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes |
title_short | Improved Lightweight Multi-Target Recognition Model for Live Streaming Scenes |
title_sort | improved lightweight multi target recognition model for live streaming scenes |
topic | model optimization object detection attention mechanism live streaming |
url | https://www.mdpi.com/2076-3417/13/18/10170 |
work_keys_str_mv | AT zongweili improvedlightweightmultitargetrecognitionmodelforlivestreamingscenes AT kaiqiao improvedlightweightmultitargetrecognitionmodelforlivestreamingscenes AT jianingchen improvedlightweightmultitargetrecognitionmodelforlivestreamingscenes AT zhenyuli improvedlightweightmultitargetrecognitionmodelforlivestreamingscenes AT yanhuizhang improvedlightweightmultitargetrecognitionmodelforlivestreamingscenes |