"To-do" have not been updated

[ayushmalpe1]

In the "3. What's in this repository" section, the (Todo) Object detection using Neural Network (TensorFlow Lite) (Todo) Object detection using YOLO v3 (RPi 4 only) have not been updated yet. Are there any plans to update them ? Is there any plan to use latest yolo model (as of Sept 2024) and include it in this repository ? Also, will this work with RasPi4/5 ?

[automaticdai]

Hi,

Thanks for your questions.

• (Todo) Object detection using Neural Network (TensorFlow Lite)
• (Todo) Object detection using YOLO v3 (RPi 4 only) has not yet been updated.
• Is there any plan to use the latest Yolo model (as of Sept 2024) and include it in this repository?

This repo is less maintained now. The current plan is to include YOLO v7. The workload for porting should be minimal and it should work out of the box. The only concern is performance as RPi is CPU only (unlike other platforms, e.g. Jetson, which has a dedicated GPU). A simple workaround is to reduce the model size and the image resolution (320x240). If we have an FPS lower than 5-10, that would simply suggest it is barely usable.

• Also, will this work with RasPi 4/5? It works with RPi 4 but It has not been tested on RPi 5 (as I haven't got one), but it should work automatically.

[ayushmalpe1]

Thanks a lot for the reply. I am sorry to ask more. Can i please get an ETA for implementation of YOLO v7 ? I shall be needing it in 5-6 days. If not, can you please send me reference links for making it work ? Thanks again.

[automaticdai]

Hi,

I am causally maintaining this repo so no ETA can be guaranteed.

The YOLO v7 that I am looking at can be found here: Yolo v7

However, this is a generic solution for performant machines. Work must be done to make it work on RPi.