Treffer: Food Object Detection Using Custom-Trained YOLOv8 with Roboflow Integration.

This study presents a complete, real-time food detection pipeline integrating the YOLOv8 object detection framework and the Roboflow dataset management platform. A dataset with a total of 1,800 high resolution images of six food categories (egg, chicken, carrot, apple, bread, and burger) was developed by using Roboflow for annotation and augmentation, and it was trained through YOLOv8 in a PyTorch (Ultralytics) framework. Strong performance was reflected in experimental results, with a mean Average Precision (mAP@0.5) of above 92% and real-time processing speed of more than 25 FPS on NVIDIA T4 GPUs. Improvements of 5% in precision and 7% in recall were observed for the proposed approach compared to SSD and YOLOv3 baselines. Thus, the findings validated that the YOLOv8-Roboflow pipeline is a flexible and effective option for modern food detection applications in health monitoring and smart kitchen systems. [ABSTRACT FROM AUTHOR]

Copyright of Engineering, Technology & Applied Science Research is the property of Engineering, Technology & Applied Science Research and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)