- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I Intel experts,
I'm really interesting with model person-detection-retail-0013 for personal detection.
I have retrained SSD-mobileNet model with coco_person dataset. It is used to dectect people on surveilance camera. However, the accuracy is not as well as your model.
So I have some question:
- How do you choose your dataset?
- Is there origin architecture of SSD-mobileNet or you have modified some where?
Thank you for any supporting from Intel!
Best Regards,
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear pham, hai,
Unfortunately we cannot disclose our data source for Intel models. But we appreciate the compliment ! Feel free to use the intel models for any of your applications.
The base of the intel models are here:
https://github.com/opencv/training_toolbox_caffe
Specifically look at the models directory.
Thanks !
Shubha
![](/skins/images/B7BA37F9CE4B52FF7CAE1FB0388839C8/responsive_peak/images/icon_anonymous_message.png)
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page