BROWSE
Detail
Conference Paper
자원 제한적인 마이크로 컨트롤러 기반 이미지 분류기 설계 및 구현
Cited - time in 
Share 
Share
- Issue Date
- 2023-06
- Citation
- 대한전자공학회 학술 대회 (하계) 2023, pp.1186-1190
- Publisher
- 대한전자공학회
- Language
- Korean
- Type
- Conference Paper
- Abstract
- This paper presents a method for optimizing image classification models to deploy them on the TensorFlow Lite micro (TFLM)[3] framework and microcontroller-based hardware. First, we get a pre-trained model by the proposed method, and next, we re-train the model based on the target dataset. Finally, we deploy the optimized model to the target hardware. Particularly, we perform image classification inference using the STM32F746G-Discovery[2] board, which is our target hardware environment. First, we get a pre-trained model by the proposed method, and next, we re-train the model based on the target dataset. Finally, we deploy the optimized model to the target hardware. Particularly, we perform image classification inference using the STM32F746G-Discovery[2] board, which is our target hardware environment. Therefore, this paper provides useful guidelines for the optimization of image classification models with the TFLM framework on microcontroller-based hardware.
- KSP Keywords
- Classification models, Image classification, Optimized model, Pre-trained model, microcontroller-based, model-based