{"title":"智能餐厅和自助咖啡厅的食物识别功能","authors":"M. Gerasimchuk, A. Uzhinskiy","doi":"10.1134/S1547477124010059","DOIUrl":null,"url":null,"abstract":"<p>In recent years, deep learning has been applied to different tasks in the food recognition field. Some promising solutions have been proposed. Due to the complexity of background food, the problem of pattern recognition on a limited dataset is still challenging. Experiments were conducted on a self-collected dataset with canteen trays, containing images of various dishes depending on the day of the week. The main objective of this work is to compare the effectiveness of modern object detection architectures, namely, YOLO_v5, YOLO_v6, YOLO_v7, and YOLO_v5, with a custom classifier. The experimental results showed that the custom classifier was needed to effectively distinguish dishes with high performance.</p>","PeriodicalId":730,"journal":{"name":"Physics of Particles and Nuclei Letters","volume":null,"pages":null},"PeriodicalIF":0.4000,"publicationDate":"2024-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Food Recognition for Smart Restaurants and Self-Service Cafes\",\"authors\":\"M. Gerasimchuk, A. Uzhinskiy\",\"doi\":\"10.1134/S1547477124010059\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>In recent years, deep learning has been applied to different tasks in the food recognition field. Some promising solutions have been proposed. Due to the complexity of background food, the problem of pattern recognition on a limited dataset is still challenging. Experiments were conducted on a self-collected dataset with canteen trays, containing images of various dishes depending on the day of the week. The main objective of this work is to compare the effectiveness of modern object detection architectures, namely, YOLO_v5, YOLO_v6, YOLO_v7, and YOLO_v5, with a custom classifier. The experimental results showed that the custom classifier was needed to effectively distinguish dishes with high performance.</p>\",\"PeriodicalId\":730,\"journal\":{\"name\":\"Physics of Particles and Nuclei Letters\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.4000,\"publicationDate\":\"2024-03-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Physics of Particles and Nuclei Letters\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://link.springer.com/article/10.1134/S1547477124010059\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"PHYSICS, PARTICLES & FIELDS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Physics of Particles and Nuclei Letters","FirstCategoryId":"1085","ListUrlMain":"https://link.springer.com/article/10.1134/S1547477124010059","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"PHYSICS, PARTICLES & FIELDS","Score":null,"Total":0}
Food Recognition for Smart Restaurants and Self-Service Cafes
In recent years, deep learning has been applied to different tasks in the food recognition field. Some promising solutions have been proposed. Due to the complexity of background food, the problem of pattern recognition on a limited dataset is still challenging. Experiments were conducted on a self-collected dataset with canteen trays, containing images of various dishes depending on the day of the week. The main objective of this work is to compare the effectiveness of modern object detection architectures, namely, YOLO_v5, YOLO_v6, YOLO_v7, and YOLO_v5, with a custom classifier. The experimental results showed that the custom classifier was needed to effectively distinguish dishes with high performance.
期刊介绍：
The journal Physics of Particles and Nuclei Letters, brief name Particles and Nuclei Letters, publishes the articles with results of the original theoretical, experimental, scientific-technical, methodological and applied research. Subject matter of articles covers: theoretical physics, elementary particle physics, relativistic nuclear physics, nuclear physics and related problems in other branches of physics, neutron physics, condensed matter physics, physics and engineering at low temperatures, physics and engineering of accelerators, physical experimental instruments and methods, physical computation experiments, applied research in these branches of physics and radiology, ecology and nuclear medicine.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
