- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I tried executing smart classroom demo in OpenVINO 2019 R1,
I've create a faces_gallery.json with create_list.py and had been successfully create.
But when executed the demo, error happened as following.
[ ERROR ] OpenCV(4.1.0-openvino) /opt/intel/openvino_2019.1.094/deployment_tools/inference_engine/samples/smart_classroom_demo/src/reid_gallery.cpp:72: error: (-215:Assertion failed) item.size() == 1 in function 'EmbeddingsGallery'
So why the error happened? Could anyone help with this kind of problem?
Thank you.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear all,
I use version 2019R1.144.
@SC Huang: I got the error when I put multiple images in gallery.
I tried to use 1 image for each identity. Then it worked.
Hope it helps.
Best Regards,
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I got the same error.
@SC Huang
Did you fix it yet?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear folks,
Can you kindly download R2019R1.1 (just released this week) and try again ? Please report your findings here.
Thanks,
Shubha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear all,
I use version 2019R1.144.
@SC Huang: I got the error when I put multiple images in gallery.
I tried to use 1 image for each identity. Then it worked.
Hope it helps.
Best Regards,
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Tom,
Thanks for your information.
I've try with one image for each identity and it works too.
Maybe multiple images will be supported in the future...
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear Tom,
Thanks for reporting your findings to the OpenVino community. It seems 1 image for each identity is required. I will file a bug because create_list.py doesn't check for multiple images, and throw an error accordingly. See the following code :
files_list = glob.glob(dir + '*.png') + glob.glob(dir + '*.jpg')
I will certainly file a bug on this !
Thanks !
Shubha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Shubha R. (Intel) wrote:Dear Tom,
Thanks for reporting your findings to the OpenVino community. It seems 1 image for each identity is required. I will file a bug because create_list.py doesn't check for multiple images, and throw an error accordingly. See the following code :
files_list = glob.glob(dir + '*.png') + glob.glob(dir + '*.jpg')I will certainly file a bug on this !
Thanks !
Shubha
Hi Shubha,
I've tried with openVINO 2019R2,
python3 create_list.py /location_to_the_gallery'
although the faces_gallery.json can be create, but it seems no items in json file.
{}
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear SC Huang,
Did you follow these instructions taken from the README.md within C:\Program Files (x86)\IntelSWTools\openvino_2019.2.242\inference_engine\demos\smart_classroom_demo ?
## Creating a Gallery for Face Recognition
To recognize faces on a frame, the demo needs a gallery of reference images. Each image should contain a tight crop of face. You can create the gallery from an arbitrary list of images:
1. Put images containing tight crops of frontal-oriented faces to a separate empty folder. Each identity could have multiple images. Name images as `id_name.0.png, id_name.1.png, ...`.
2. Run the `create_list.py <path_to_folder_with_images>` command to get a list of files and identities in `.json` format.
In your command python3 create_list.py /location_to_the_gallery it should be python3 create_list.py <full_path>/location_to_the_gallery . Or maybe you need a dot in front of your slash like : python3 create_list.py ./location_to_the_gallery
Let me know if that helps.
Thanks !
Shubha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Shubha R. (Intel) wrote:In your command python3 create_list.py /location_to_the_gallery it should be python3 create_list.py <full_path>/location_to_the_gallery . Or maybe you need a dot in front of your slash like : python3 create_list.py ./location_to_the_gallery
Let me know if that helps.
Thanks !
Shubha
Hi Shubha,
Sorry that i simplify the location, actually my command location is full path with
python3 create_list.py $HOME/Picture/TestImage/
inside that location
ls $HOME/Picture/TestImage/
FCT.0.png FCT.1.png FCT.2.png FCT.3.png FCT.4.png FCT.5.png SCH.0.png SCH.1.png SCH.2.png SCH.3.png SCH.4.png SCH.5.png
I think I've totally follow the instruction from README.md, did I miss anything?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
I have the same problem on openvino R2.
Only "{}" on faces_gallery.json after running create_list.py
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Dear SC Huang
Please attach as a *.zip file your images and I will reproduce.
Thanks for your cooperation !
Shubha
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
why don't we just fix the create_list.py script by introducing a slash infromnt of the png and jpg pattent strings ?
so instead of :
files_list = glob.glob(dir + '*.png') + glob.glob(dir + '*.jpg')
have this:
files_list = glob.glob(dir + '/*.png') + glob.glob(dir + '/*.jpg')
without this any one who gives the gallery folder as c:\abc\pqr or /abc/pqr would get an empty json
![](/skins/images/D2683F18326913BBA0436CB7114DD569/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