Milk-V Duo | Camera + TPU + RTSP?

Hi! I’ve seen the TPU YOLOv5 example and the camera RTSP stream example. I’m wondering, is it possible to combine the two and simultaneously stream RTSP video from the CSI-2 camera while performing image classification? How to approach that? Thanks :slight_smile:

yes, it is possible to do that. Actually, many business customers have already done that. We will make a demo later.

1 Like

We will release a demo in 2 weeks.

1 Like

Thanks a lot! Can’t wait for the demo :smiley:

Hello :slight_smile: Has the demo been released yet?

Hi, any news with that? :smiley:

will be finished next week

1 Like

Hi there @Sandor :smiley: Is there any chance of the demo being released this week? :slight_smile:

@Sandor Any news? I’m really looking forward to this example

1 Like

The demo will come next week.Sorry for later~
I will reply when I put the demo in the github.

Thank you! I hope it arrives soon :slight_smile:

You can refer:

Dear @chenli.zhang thank you for great tutorial on face detection from camera!
As I understand the tutorial code is in cvitek-tdl-sdk-cv180x/sample/cvi_tdl/sample_vi_fd.c at main · milkv-duo/cvitek-tdl-sdk-cv180x · GitHub
I examine it and I cannot understand one place. Could you please explain what exactly happens in the line 205…209, line by line? Thank you very much!

  stMWConfig.stVPSSPoolConfig.stVpssMode.aenInput[0] = VPSS_INPUT_MEM;
  stMWConfig.stVPSSPoolConfig.stVpssMode.enMode = VPSS_MODE_DUAL;
  stMWConfig.stVPSSPoolConfig.stVpssMode.ViPipe[0] = 0;
  stMWConfig.stVPSSPoolConfig.stVpssMode.aenInput[1] = VPSS_INPUT_ISP;
  stMWConfig.stVPSSPoolConfig.stVpssMode.ViPipe[1] = 0;