- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
____When I run the fogbuz ncappzoo demo "street_cam", I meet AttributeError just as follow. The make is no problem. ****
root@ubuntu:~/workspace/ncsdk/ncappzoo/apps# cd street_cam
root@ubuntu:~/workspace/ncsdk/ncappzoo/apps/street_cam# ls
install-opencv-from_source.sh README.md street_cam.py
Makefile screen_shot.jpg
root@ubuntu:~/workspace/ncsdk/ncappzoo/apps/street_cam# make
making prereqs
making videos
wget -c --no-cache -P . https://raw.githubusercontent.com/nealvis/media/master/traffic_vid/licenses.txt; wget -c --no-cache -P . https://raw.githubusercontent.com/nealvis/media/master/traffic_vid/bus_station_6094_960x540.mp4; wget -c --no-cache -P .
………………….
making compile
mvNCCompile -w bvlc_googlenet.caffemodel -s 12 deploy.prototxt
mvNCCompile v02.00, Copyright @ Movidius Ltd 2016
Layer inception_3b/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_3b/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4a/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4a/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4b/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4b/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4c/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4c/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4d/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4d/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4e/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4e/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_5a/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_5a/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_5b/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_5b/pool_proj forced to im2col_v2, because its output is used in concat
/usr/local/bin/ncsdk/Controllers/FileIO.py:52: UserWarning: You are using a large type. Consider reducing your data sizes for best performance
"Consider reducing your data sizes for best performance\033[0m")
make[1]: Leaving directory '/root/workspace/ncsdk/ncappzoo/caffe/GoogLeNet'
making tiny_yolo
(cd ../../caffe/TinyYolo; make compile; cd ../../apps/street_cam; cp ../../caffe/TinyYolo/graph ./yolo_tiny.graph;)
make[1]: Entering directory '/root/workspace/ncsdk/ncappzoo/caffe/TinyYolo'
…………………………
making compile
mvNCCompile -o graph -w tiny-yolo-v1_53000.caffemodel -s 12 tiny-yolo-v1.prototxt
mvNCCompile v02.00, Copyright @ Movidius Ltd 2016
/usr/local/bin/ncsdk/Controllers/FileIO.py:52: UserWarning: You are using a large type. Consider reducing your data sizes for best performance
"Consider reducing your data sizes for best performance\033[0m")
make[1]: Leaving directory '/root/workspace/ncsdk/ncappzoo/caffe/TinyYolo'
making compile
root@ubuntu:~/workspace/ncsdk/ncappzoo/apps/street_cam# make run
making prereqs
making videos
……………………..
The file is already fully retrieved; nothing to do.
making ilsvrc12
(cd ../../data/ilsvrc12; make)
make[1]: Entering directory '/root/workspace/ncsdk/ncappzoo/data/ilsvrc12'
make[1]: Leaving directory '/root/workspace/ncsdk/ncappzoo/data/ilsvrc12'
making googlenet
(cd ../../caffe/GoogLeNet; make compile; cd ../../apps/street_cam; cp ../../caffe/GoogLeNet/graph ./googlenet.graph;)
make[1]: Entering directory '/root/workspace/ncsdk/ncappzoo/caffe/GoogLeNet'
making prereqs
(cd ../../data/ilsvrc12; make)
make[2]: Entering directory '/root/workspace/ncsdk/ncappzoo/data/ilsvrc12'
make[2]: Leaving directory '/root/workspace/ncsdk/ncappzoo/data/ilsvrc12'
making prototxt
Prototxt file already exists
making caffemodel
caffemodel file already exists
making compile
mvNCCompile -w bvlc_googlenet.caffemodel -s 12 deploy.prototxt
mvNCCompile v02.00, Copyright @ Movidius Ltd 2016
Layer inception_3b/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_3b/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4a/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4a/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4b/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4b/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4c/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4c/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4d/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4d/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_4e/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_4e/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_5a/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_5a/pool_proj forced to im2col_v2, because its output is used in concat
Layer inception_5b/1x1 forced to im2col_v2, because its output is used in concat
Layer inception_5b/pool_proj forced to im2col_v2, because its output is used in concat
/usr/local/bin/ncsdk/Controllers/FileIO.py:52: UserWarning: You are using a large type. Consider reducing your data sizes for best performance
"Consider reducing your data sizes for best performance\033[0m")
make[1]: Leaving directory '/root/workspace/ncsdk/ncappzoo/caffe/GoogLeNet'
making tiny_yolo
(cd ../../caffe/TinyYolo; make compile; cd ../../apps/street_cam; cp ../../caffe/TinyYolo/graph ./yolo_tiny.graph;)
make[1]: Entering directory '/root/workspace/ncsdk/ncappzoo/caffe/TinyYolo'
making prototxt
Prototxt file already exists
making caffemodel
caffemodel file already exists
making compile
mvNCCompile -o graph -w tiny-yolo-v1_53000.caffemodel -s 12 tiny-yolo-v1.prototxt
mvNCCompile v02.00, Copyright @ Movidius Ltd 2016
/usr/local/bin/ncsdk/Controllers/FileIO.py:52: UserWarning: You are using a large type. Consider reducing your data sizes for best performance
"Consider reducing your data sizes for best performance\033[0m")
make[1]: Leaving directory '/root/workspace/ncsdk/ncappzoo/caffe/TinyYolo'
making compile
making run
python3 ./street_cam.py
Traceback (most recent call last):
File "./street_cam.py", line 9, in
import mvnc_simple_api as mvnc
File "../../ncapi2_shim/mvnc_simple_api.py", line 10, in
class mvncStatus(Enum):
File "../../ncapi2_shim/mvnc_simple_api.py", line 23, in mvncStatus
NOT_ALLOCATED = mvncapi2.Status.NOT_ALLOCATED
File "/usr/local/lib/python3.5/dist-packages/mvnc/mvncapi.py", line 44, in getattr
return getattr(self.new_target, attr)
File "/usr/lib/python3.5/enum.py", line 274, in getattr
raise AttributeError(name) from None
AttributeError: NOT_ALLOCATED
Makefile:64: recipe for target 'run' failed
make: *** [run] Error 1
root@ubuntu:~/workspace/ncsdk/ncappzoo/apps/street_cam#
- Tags:
- OpenCV
Link Copied
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @ideallyworld ,
Thanks for contacting us. I have tested the street_cam examples on a Ubuntu 16.04 host, at first I encountered an error as I was only using one NCS device. To run this example you may actually need two NCS devices. May I ask which version of the NCSDK are you using? Please make sure you have downloaded the corresponding NCAPPZOO version to the NCSDK version you are using. Also are you running on a VM? If you can provide any details of your environment and setup that could help me reproduce the problem, it would be great.
Best Regards,
@Luis_at_Intel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Luis_at_Intel ,
Thank you very much for your quickly response.
1st:I was only using one NCS device. And then I run W/ another demo "workspace/ncsdk/ncappzoo/apps/video_objects", only one NCS device is enough for this demo, But it got the same make run error just as follow:
………………………..
3|:443… connected.
HTTP request sent, awaiting response… 416 Range Not Satisfiable
The file is already fully retrieved; nothing to do.
making run_py
python3 ./video_objects.py
Traceback (most recent call last):
File "./video_objects.py", line 8, in
import mvnc_simple_api as mvnc
File "../../ncapi2_shim/mvnc_simple_api.py", line 10, in
class mvncStatus(Enum):
File "../../ncapi2_shim/mvnc_simple_api.py", line 23, in mvncStatus
NOT_ALLOCATED = mvncapi2.Status.NOT_ALLOCATED
File "/usr/local/lib/python3.5/dist-packages/mvnc/mvncapi.py", line 44, in getattr
return getattr(self.new_target, attr)
File "/usr/lib/python3.5/enum.py", line 274, in getattr
raise AttributeError(name) from None
AttributeError: NOT_ALLOCATED
Makefile:41: recipe for target 'run_py' failed
make: *** [run_py] Error 1
…………………………………..
2nd: NCSDK-1.12.01.01.tar.gz , VMware® Workstation 14 Pro version:14.1.5 build-10950780.
Thanks again.
BRs,
@ideallyworld
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @ideallyworld ,
Thanks for the information. I've tested the NCSDK-1.12.01.01 on a VMware(R) Worksation 15 Player 15.0.0 build-10134415 running Ubuntu 16.04 and had no issues testing the street_cam sample. One thing I noticed though by looking at the logs you provided is the following line "File "../../ncapi2_shim/mvnc_simple_api.py", line 10, in ". Doing a comparison between NCAPPZOO versions, this ncap2_shim is found in NCAPPZOOv2 and not found in NCAPPZOOv1.The ncappzoo/ncapi2_shim project allows NCAPIv1 python code to run with NCAPIv2, but I don't think its backwards compatible (meaning NCAPIv2 python code to run with NCAPIv1). I have tested this (NCAPPZOOv2 samples on NCAPIv1) and I get the same error as you. I think the solution here is to remove and reinstall the correct NCAPPZOO version for your NCSDK-1.12.01.01, this should definitely resolve your issue. Hope this helps!
Cheers,
@Luis_at_Intel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Luis_at_Intel ,
Thank your very much for your comments. I reinstall the NCAPPZOO version W/(movidius/ncappzoo: master branch Updated Dec 16, 2018 by squeakus). And the demo "workspace/ncsdk/ncappzoo/apps/video_objects" is running OK. So you are all right. The root cause is the NCAPPZOO version is not right W/ NCSDK-1.12.01.01.
Then I try another demo "video_face_matcher" , It's running OK,and I got the camera video. But it only could running few seconds and then closed with follow:
Total Difference is: 1.88481992483
FAIL! File camera frame 168 does not match ./validated_images/valid.jpg
Total Difference is: 1.93335545063
FAIL! File camera frame 169 does not match ./validated_images/valid.jpg
Total Difference is: 1.88070523739
FAIL! File camera frame 170 does not match ./validated_images/valid.jpg
Total Difference is: 1.90008598566
FAIL! File camera frame 171 does not match ./validated_images/valid.jpg
Total Difference is: 1.87470090389
FAIL! File camera frame 172 does not match ./validated_images/valid.jpg
Total Difference is: 1.95423680544
FAIL! File camera frame 173 does not match ./validated_images/valid.jpg
Total Difference is: 1.88118928671
FAIL! File camera frame 174 does not match ./validated_images/valid.jpg
Total Difference is: 1.90649461746
FAIL! File camera frame 175 does not match ./validated_images/valid.jpg
select timeout
VIDIOC_DQBUF: Resource temporarily unavailable
No image from camera, exiting
root@ubuntu:~/workspace/ncsdk/ncappzoo-master/apps/video_face_matcher#
I don‘t know why could be this.
Thanks again.
BRs,
@ideallyworld
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @ideallyworld ,
Thanks for confirming your issue was resolved, I am glad the suggestions worked.
As far as your other question goes with the
"video_face_matcher"
sample, the FAIL! is expected as it is checking your webcam feed and comparing the frames to the valid.jpg image located in the ~/ncappzoo/apps/video_face_matcher/validated_images
directory. I took a picture of me using my webcam application and replaced the valid.jpg image with the resulting .jpg file, then I re-ran the sample and saw the FAIL! convert to PASS! as the new image matches the camera feed frames. Just wanted to clarify that to you for this particular video_face_matcher sample, I hope my explanation makes sense.Now, I think the error
<i>VIDIOC_DQBUF: Resource temporarily unavailable No image from camera, exiting</i>
is related to the webcam, may I ask if you can share your webcam specs, is a built-in or a USB webcam, etc..Regards,
@Luis_at_Intel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Luis_at_Intel ,
Thank you very much for your explains. I got it. I will try replaced the valid.jpg W/ my webcamera images.
Additionally. My camera is a (USB 2.0 ) webcam. 640 x 480 pixels , 30fps W/ YUV format or 15fps W/ MJPEG format , support linux/ubuntu/openwrt/Pi .
Thanks again.
BRs,
@ideallyworld
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @ideallyworld ,
Thanks for the information, the closest camera I have to your configuration is a USB2.0, 720p @ 30 FPS. The program runs just fine and there is no sign of the problem you experienced, is there a way for you to try and test using a different webcam? I hope you do and this resolves your problem, let me know.
Regards,
@Luis_at_Intel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Luis_at_Intel ,
Thanks for your reply. Yeah , I spend some days to buy a new USB3.0 1080p @50 FPS webcam on the Internet and retry this demo. It's seems better and could running much more time. But it still will
closed as follow:
…..
Total Difference is: 1.25864952803
FAIL! File camera frame 3574 does not match ./validated_images/valid.jpg
Total Difference is: 1.23989248276
FAIL! File camera frame 3575 does not match ./validated_images/valid.jpg
Total Difference is: 1.23096191883
FAIL! File camera frame 3576 does not match ./validated_images/valid.jpg
Total Difference is: 1.25267809629
FAIL! File camera frame 3577 does not match ./validated_images/valid.jpg
Total Difference is: 1.25684344769
FAIL! File camera frame 3578 does not match ./validated_images/valid.jpg
Total Difference is: 1.2717924118
FAIL! File camera frame 3579 does not match ./validated_images/valid.jpg
Total Difference is: 1.27381694317
FAIL! File camera frame 3580 does not match ./validated_images/valid.jpg
select timeout
VIDIOC_DQBUF: Resource temporarily unavailable
No image from camera, exiting
root@ubuntu:~/workspace/ncsdk/ncappzoo-master/apps/video_face_matcher#
Is this normal phenomenon?
Additionally, I have some questions about how to improve the face matcher accuracy ? Could you please give me some methods or suggestions ?
Thank you very much?
BRs,
@ideallyworld
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Luis_at_Intel ,
Thanks for your reply. Yeah , I spend some days to buy a new USB3.0 1080p @50 FPS webcam on the Internet and retry this demo. It's seems better and could running much more time. But it still will
closed as follow:
…..
Total Difference is: 1.25864952803
FAIL! File camera frame 3574 does not match ./validated_images/valid.jpg
Total Difference is: 1.23989248276
FAIL! File camera frame 3575 does not match ./validated_images/valid.jpg
Total Difference is: 1.23096191883
FAIL! File camera frame 3576 does not match ./validated_images/valid.jpg
Total Difference is: 1.25267809629
FAIL! File camera frame 3577 does not match ./validated_images/valid.jpg
Total Difference is: 1.25684344769
FAIL! File camera frame 3578 does not match ./validated_images/valid.jpg
Total Difference is: 1.2717924118
FAIL! File camera frame 3579 does not match ./validated_images/valid.jpg
Total Difference is: 1.27381694317
FAIL! File camera frame 3580 does not match ./validated_images/valid.jpg
select timeout
VIDIOC_DQBUF: Resource temporarily unavailable
No image from camera, exiting
root@ubuntu:~/workspace/ncsdk/ncappzoo-master/apps/video_face_matcher#
Is this normal phenomenon?
Additionally, I have some questions about how to improve the face matcher accuracy ? Could you please give me some methods or suggestions ?
Thank you very much?
BRs,
@ideallyworld
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @ideallyworld ,
I am not sure what could be causing the problem VIDIOC_DQBUF: Resource temporarily unavailable No image from camera, exiting, looks like maybe the webcam is getting reset after a certain amount of time. May I ask after how many minutes the problem occurs? I'd like to see if I get the same problem.
I don't have any suggestions on how to improve face matcher accuracy, you could take a look at the following resources if you want to learn more and get ideas on how to go about your question:
- https://movidius.github.io/blog/deploying-custom-caffe-models
- https://movidius.github.io/blog/mvNCCheck
- https://movidius.github.io/ncsdk/configure_network.html
Regards,
@Luis_at_Intel
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi @Luis_at_Intel ,
Thank you very much for reply and suggestions. It's about 4 minutes.
Let me give more detail about the question of face_matcher accuracy. I just want to try the Demo"face_matcher accuracy" just like the demo "dogsvscats". So could I get the resource of the "face_matcher accuracy" just like the demo "dogsvscats"? The Learning network is included which I could tranning it and improve it.
Thanks again.
BRs,
@ideallyworld
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page