Ncsdk libraries installation

[idata]

After successfully installing ncsdk in Raspberry pi, should I need to install opencv, caffe etc using "make example" command or is it enough to install opencv, tensorflow etc using pip command inside another directory other than ncsdk directory.

Thanks in advance

Hashir

[idata]

Hi @Hashir

Did you install the API-only version of the NCSDK or the full version?

Sincerely,

Sahira

[idata]

@Sahira_at_Intel ,

Am trying to install API version from api/src directory, but I got following error for both ncsdk version 2 and version 1.Here I'm attaching the error I got. Pls help me to solve this.

Thanks in advance

[idata]

Hi @Hashir

I don't think the error log was attached!

Best Regards,

Sahira

[idata]

@Sahira_at_Intel , thanks. I solved that issue.

But after I ran tiny yolo real time object recognition, my webcam shows real-time video without any detection or bounding boxes, why ?

Pls help me out

[idata]

Hi @Hashir

Thank you for confirming your installation issue is resolved. Please open a separate thread regarding the object detection issue with additional details (NCSDK Version, VM?, model).

Regards,

Jesus

[idata]

@Jesus_at_Intel , actually am following realtime object recognition using tinyYolo done by @PINTO (GitHub). My ncsdk version is version 1 but just API version only not full version. After running tinyYoLO model in my Raspberry pi 3 using movidius, my webcam shows real time live video only not a detection or even without a single bounding box or prediction. Pls help me to get through this.

Thanks in advance

Hashir

[idata]

Hi @Hashir

I can see that PINTO is helping you out on your other thread: https://ncsforum.movidius.com/discussion/985/attributeerror-module-mvnc-mvncapi-has-no-attribute-global-set-option#latest

I will go ahead and close this thread, feel free to open a new one if you have additional questions.

Regards,

Jesus