[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2153":3},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":18,"tags":20,"view_count":15,"doi":24,"paper":25,"created_at":40},2153,"Design and Deployment of Multi-Class Image Classification Model on ESP32 Using a Lightweight CNN Approach","https:\u002F\u002Fdoi.org\u002F10.57041\u002Fp6f08552","Tremendous progress has been made in developing intelligent and autonomous systems in recent years, driven by AI software tools such as Chatbots and large language models (LLMs) for various applications. This has led to the adoption of smart technology solutions. Nonetheless, computer vision requires significant computational power, making it difficult to run on devices such as the ESP32. To showcase the development of smart technologies leveraging controller platforms, the research conducted training and deployment of an optimal multi-class image classification model on the ESP32 microcontroller platform, ensuring acceptable predictive performance across four classes: person, fruit, car, and unknown. It is important to clearly distinguish between the proposed system, which performs image classification and object detection. Regarding model training, the images were preprocessed to minimise computational cost. Nonetheless, the trained model achieved a peak validation accuracy of 99.34±1% when evaluated using five-fold cross-validation, with an average validation accuracy of 98±1% without altering the model or the training process, producing similar results. Moreover, TFLite Quantisation, a variant of the INT8 quantisation technique, was used to optimise the model size. The model was converted to C array format while retaining most features. To improve accuracy during training, the training dataset was augmented with large, high-quality images that had been filtered out. The chosen hardware used for this project was the ESP32-S3 N16R8 microcontroller with built-in OV2640 camera and the 3.2'' TFT to visualize the process of computer vision task execution in real time. The model works smoothly and predicts accurately, even with just a few samples per frame, to decrease response time latency. Successfully running the multi-class image classification model on the ESP32 microcontroller is the evidence of full independence of embedded vision solution without having anything to do with cloud technology.","近年来，在聊天机器人（Chatbots）和大语言模型（LLMs）等AI软件工具的推动下，智能自主系统的开发取得了巨大进展，并被应用于各种场景。这促进了智能技术解决方案的采用。然而，计算机视觉需要大量的计算能力，使其难以在ESP32等设备上运行。为了展示利用控制器平台开发智能技术的可行性，本研究在ESP32微控制器平台上训练并部署了一个最优多类图像分类模型，确保在四个类别（人、水果、汽车和未知）上具有可接受的预测性能。需要明确区分所提出的系统与执行图像分类和目标检测的系统。在模型训练方面，对图像进行了预处理以最小化计算成本。尽管如此，使用五折交叉验证评估时，训练模型达到了99.34±1%的最高验证准确率，在未改变模型或训练过程的情况下，平均验证准确率为98±1%，结果相似。此外，采用TFLite量化（TFLite Quantisation），即INT8量化技术的一种变体，来优化模型大小。模型被转换为C数组格式，同时保留了大部分特征。为了提高训练期间的准确率，训练数据集补充了此前被过滤掉的大尺寸高质量图像。本项目选用的硬件为内置OV2640摄像头的ESP32-S3 N16R8微控制器，以及3.2英寸TFT显示屏，用于实时可视化计算机视觉任务的执行过程。该模型运行流畅、预测准确，即使每帧仅有少量样本也能正常工作，从而降低响应时间延迟。在ESP32微控制器上成功运行多类图像分类模型，证明了嵌入式视觉解决方案的完全独立性，无需依赖云技术。",null,"International Journal of Emerging Engineering and Technology","2026-09-09T00:00:00Z","论文",10,false,0,{"impact":15,"substance":15,"depth":15,"authority":15,"freshness":15,"relevant":15,"comment":17},"该论文聚焦ESP32微控制器上的轻量级CNN图像分类部署，属嵌入式计算机视觉通用技术研究，未涉及三农、农业信息化或智慧农业应用场景，与平台主题不相关。",[19],{"name":10,"url":6},[21,22,23],"农业人工智能","边缘计算","图像识别","10.57041\u002Fp6f08552",{"doi":24,"openalex_id":26,"authors":27,"venue":10,"cited_by_count":15,"oa_url":32,"card":33,"direction":37,"ingested_from":39},"W7212075138",[28,30],{"name":29,"orcid":9},"Abdul Hannan",{"name":31,"orcid":9},"Rao M. Asif","https:\u002F\u002Fgrsh.org\u002Fjournal1\u002Findex.php\u002Fijeet\u002Farticle\u002Fdownload\u002F202\u002F114",{"tldr":34,"method":35,"finding":36,"direction":37,"opportunity":38},"在ESP32微控制器上训练并部署轻量CNN多类图像分类模型，实现无云端的嵌入式视觉。","轻量CNN、TFLite INT8量化、五折交叉验证、ESP32-S3与OV26","模型验证准确率达99.34±1%，量化后仍保持性能，可在ESP32上实时准确分类。","智慧农业 \u002F 农业物联网","可探索将此类轻量模型用于田间作物\u002F病虫害实时识别，解决农业边缘设备算力与网络限制。","openalex","2026-09-11T23:30:14.512845Z"]