# Edge AI deployment in practice: running object detection with TFLite Micro on STM32F407

- Source: 古月居 · 微信公众号
- Published: 2026-10-20T09:20:00.000Z
- RoboHot: https://robohot.news/en/items/yo4mxds4us1ziom6e2cv12e9n
- Original: https://mp.weixin.qq.com/s?__biz=MzU1NjEwMTY0Mw==&mid=2247622253&idx=1&sn=089f8f029d920752cbb33cf7065e27b6&chksm=faec842fe4755869498e13db038d31ad29037b8c050423ed32c66a9f390b25b003db7b3baa4b&scene=126&sessionid=1792506638#rd

## Summary

Using the STM32F407 (168MHz, 192KB SRAM) as an example, the article walks through the full chain for object detection with TFLite Micro: model conversion, full int8 integer quantization, porting, and inference.