How to create a new application

From the docs (http://doc.kneron.com/docs/#getting_start/#8-create-new-sdk-application) I can see that I should copy an application template from "scpu/project/tiny_yolo_v3".

Where can I find the folder "scpu"?

Thanks.

Comments

  • Kneron provided firmware sdk for all KL520 device. Please download SDK from Development Center (https://www.kneron.com/tw/support/developers/) or get sdk v1.3 from following link.

    You will find the folder "scpu" in SDK.

  • Thanks. I am trying to use multiple models in a single application and I can see from the docs that an example is already doing it:

    However, I cannot find such example (age_gender) inside the "scpu/project" folder (there is only the "tiny_yolo_v3" foider).

    Where can I find it?

    Thank you.

  • For business policy, we no longer provide full application code of "age_gender" for free. But you can still find the model binaries in host_lib example (\host_lib_v0.9.0-early\app_binaries\KL520\age_gender).

    (Download host_lib_v0.9.0-early.zip from Kneron Developers)

    For the multi-models purpose, please combine your models to single binary by toolchain, and use api "kdp_dme_configure" to config each model id so you can run "kdp_dme_inference" with the model id which you hope to perform.

  • edited January 2021

    I see. However I am looking for an example of a ISI implementation of multi-model, not a dme one.

    I suggest you to specify inside the documentation that the age_gender code is no longer available because it is not clear from what I saw.

    Thank you very much for you help.

  • Thanks for your suggestion.

    We will make modification to it.

  • Hello Ethon,

    I managed to make the "age_gender" model work in ISI mode. However, when multiple faces are visible in a input image, only one face is detected.

    My question is: is it correct to have only one output bounding box from the program?

    I would like to detect all the faces but when I pass the struct "fd_age_gender_s" to "kdp_isi_retrieve_res", count = 1 but it should be count = 3.

    Thank you.

  • Hi Paolo,

    Yes, now the App. age_gender only returns result with the face which is the highest score.

    If you just want to detect where the faces are, please refer to the example ssd_fd and you can get result with all the faces.

  • Ok, thank you.

The discussion has been closed due to inactivity. To continue with the topic, please feel free to post a new discussion.