Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.

Multil Label tracking for Movidius

idata
Employee
776 Views

Hi,

 

I just played around with the Movidius and it is amazing how the SDK make our life easy to enable machine learning features :smile:

 

I have a technical question which i am stuck for my project. I wish to detect in multiple object using finetune technique just like what i have attached here. Could I get advice which network is the best to achieve this?

 

I tried fine-tuning with GoogLeNet (classification) to identify a robber with mask & gun. My network will eventually converged to just 2 output. First one, normal person; while second is robber with mask & gun. Trained accuracy is 99%. It is not able to differentiate between normal person and robber.

 

Appreciate your advice. Thank you.

0 Kudos
1 Reply
Reply