多元融合YOLO:通过多源信息融合重新定义葡萄酒品种识别
Jialiang Peng1, Cheng Ouyang1, Hao Peng1
1College of Information and Intelligence, Hunan Agricultural University, Changsha 410128, China.
这项研究介绍了MultiFuseYOLO,这是一种用于葡萄酒品种识别的新型深度学习模型. 它通过融合多来源信息,显著提高了准确性,特别是在视觉上相似的葡萄中.
科学领域:
- 农业科学 农业科学
- 计算机视觉 计算机视觉
- 机器学习 机器学习
背景情况:
- 传统的深度学习模型由于品种之间的高度相似性,难以识别葡萄酒品种.
- 单一特征分类方法不足以准确区分.
研究的目的:
- 开发一种多源信息融合方法,以提高葡萄酒品种的识别能力.
- 提高识别视觉上相似的葡萄品种的准确性和可靠性.
主要方法:
- 优化YOLOV7模型以创建WineYOLO-RAFusion,以改善水果定位和识别.
- 将多源信息融合集成到WineYOLO-RAFusion中,从而产生了MultiFuseYOLO模型.
- 使用SynthDiscrim算法作为多源信息融合的核心组件.
主要成果:
- MultiFuseYOLO显著优于现有的模型,其精度,回忆和F1分数分别为0.854,0.815和0.833.
- 区分查多内和索维尼昂布朗品种的精度大幅增加,分别从0.512增加到0.813和0.533增加到0.775.
- 该模型在识别具有挑战性的,视觉相似的葡萄酒品种方面表现出卓越的表现.
结论:
- 多聚合YOLO模型为葡萄酒品种识别提供了强大的解决方案.
- 多源信息融合对于实现高精度识别至关重要,特别是对于相似品种.
- 这种方法提高了葡萄栽培中自动葡萄识别系统的可靠性.
更多相关视频
07:34Author Spotlight: Exploring the Fermentation Microbiome Through Next-Generation Sequencing
Published on: December 1, 2023
10:40Comprehensive Workflow for the Genome-wide Identification and Expression Meta-analysis of the ATL E3 Ubiquitin Ligase Gene Family in Grapevine
Published on: December 22, 2017
相关概念视频
Tagging and Fusion Proteins
Multi-input and Multi-variable systems
In the absence...
Multi-species Conserved Sequences
Although the genome of each species varies greatly from each other, a few sequences are highly conserved. Such conserved...
FISH - Fluorescent In-situ Hybridization
Aggregates Classification
Petrographic classification groups aggregates based on common mineralogical characteristics. Some of the common mineral groups found in aggregates are...
Classification of Titrimetric Analysis Based on Reaction Types
Titrations between an acid and a base lead to neutralization reactions that form...
