OpenVINO for NCSDK 1, instead of NCSDK 2?

[PJIA01]

Hi, all, I am testing NCSDK 1 with a Raspberry Pi 4. I'm using OpenVINO 2019_R3.  And I've been strictly following https://docs.openvinotoolkit.org/latest/_docs_install_guides_installing_openvino_raspbian.html

But I still obtained the following ERROR messages. I'm guessing the built example is for NCSDK 2, instead of NCSDK 1???

pi@raspberrypi:/opt/intel/openvino/inference_engine/samples/build $ ./armv7l/Release/object_detection_sample_ssd -m face-detection-adas-0001.xml -d MYRIAD -i cam.jpg
[ INFO ] InferenceEngine: 
    API version ............ 2.1
    Build .................. custom_releases/2019/R3_cb6cad9663aea3d282e0e8b3e0bf359df665d5d0
    Description ....... API
Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ]     cam.jpg
[ INFO ] Loading Inference Engine
[ INFO ] Device info: 
    MYRIAD
    myriadPlugin version ......... 2.1
    Build ........... 30677
[ INFO ] Loading network files:
    face-detection-adas-0001.xml
    face-detection-adas-0001.bin
[ INFO ] Preparing input blobs
[ INFO ] Batch size is 1
[ INFO ] Preparing output blobs
terminate called after throwing an instance of '__gnu_cxx::__concurrence_lock_error'
  what():  __gnu_cxx::__concurrence_lock_error
Aborted
 

Anyway, can anybody give me a hand?

Cheers & Merry Christmas...

Pei

[David_C_Intel]

Hello Pei,

Thank you for reaching out. 

Could you please answer the following:

- Which Raspbian OS version are you using?

- Are you running the raspbery pi with a fresh image?

- Which CMake version is installed? (type "cmake --version")

Regards,

David

[PJIA01]

Hey, David.

Happy new year. Thank you for your prompt reply.

pi@raspberrypi:~ $ lsb_release -a
No LSB modules are available.
Distributor ID:    Raspbian
Description:    Raspbian GNU/Linux 10 (buster)
Release:    10
Codename:    buster
pi@raspberrypi:~ $ cmake --version
cmake version 3.13.4

CMake suite maintained and supported by Kitware (kitware.com/cmake).

I update&upgrade a lot of packages already. It's NOT a fresh one.

Cheers

Pei

[David_C_Intel]

Hello Pei,

Happy holidays for you as well. 

We tested it from our end and it worked as expected. Try to do it with a fresh SD card image and tell us if it works.
These are the commands used to run the sample:

In a custom directory use:

 mkdir build && cd build
 
cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_FLAGS="-march=armv7-a" /opt/intel/openvino/deployment_tools/inference_engine/samples

make -j2 object_detection_sample_ssd

Download the model files:

 
 wget --no-check-certificate https://download.01.org/opencv/2019/open_model_zoo/R1/models_bin/face-detection-adas-0001/FP16/face-detection-adas-0001.bin

 wget --no-check-certificate https://download.01.org/opencv/2019/open_model_zoo/R1/models_bin/face-detection-adas-0001/FP16/face-detection-adas-0001.xml

Run the sample with the specified input image:

 ./armv7l/Release/object_detection_sample_ssd -m face-detection-adas-0001.xml -d MYRIAD -i <path_to_image>

Regards,

David

[crook__russell]

I am also using NCSDK 1.  I have successfully demonstrated to face detection demo but have not been able to get the age-gender-recognition demo to work.  

This is the command line that I used for the face detection

$  ~/build/armv7l/Release/object_detection_sample_ssd -m ~/build/face-detection-adas-0001.xml -d MYRIAD -i ~/Pictures/_DSC0523.jpg

This appears to work fine.

[ INFO ] InferenceEngine:
        API version ............ 2.1
        Build .................. custom_releases/2019/R3_cb6cad9663aea3d282e0e8b3e0bf359df665d5d0
        Description ....... API
Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ]     /home/pi/Pictures/_DSC0523.jpg
[ INFO ] Loading Inference Engine
[ INFO ] Device info:
        MYRIAD
        myriadPlugin version ......... 2.1
        Build ........... 30677
[ INFO ] Loading network files:
        /home/pi/build/face-detection-adas-0001.xml
        /home/pi/build/face-detection-adas-0001.bin
[ INFO ] Preparing input blobs
[ INFO ] Batch size is 1
[ INFO ] Preparing output blobs
[ INFO ] Loading model to the device
[ INFO ] Create infer request
[ WARNING ] Image is resized from (1338, 893) to (672, 384)
[ INFO ] Batch size is 1
[ INFO ] Start inference
[ INFO ] Processing output blobs
[0,1] element, prob = 1    (715,175)-(772,249) batch id : 0 WILL BE PRINTED!
[1,1] element, prob = 1    (601,201)-(658,276) batch id : 0 WILL BE PRINTED!
[2,1] element, prob = 1    (497,227)-(553,302) batch id : 0 WILL BE PRINTED!
[3,1] element, prob = 1    (530,473)-(592,557) batch id : 0 WILL BE PRINTED!
[4,1] element, prob = 1    (1125,129)-(1199,211) batch id : 0 WILL BE PRINTED!
[5,1] element, prob = 1    (378,153)-(439,234) batch id : 0 WILL BE PRINTED!
[6,1] element, prob = 1    (940,181)-(1008,267) batch id : 0 WILL BE PRINTED!
[7,1] element, prob = 1    (1215,265)-(1284,359) batch id : 0 WILL BE PRINTED!
[8,1] element, prob = 1    (999,486)-(1068,574) batch id : 0 WILL BE PRINTED!
[9,1] element, prob = 0.999512    (227,102)-(291,185) batch id : 0 WILL BE PRINTED!
[10,1] element, prob = 0.999512    (868,481)-(929,564) batch id : 0 WILL BE PRINTED!
[11,1] element, prob = 0.999023    (123,134)-(198,229) batch id : 0 WILL BE PRINTED!
[12,1] element, prob = 0.998535    (825,218)-(883,292) batch id : 0 WILL BE PRINTED!
[13,1] element, prob = 0.995605    (1042,185)-(1083,238) batch id : 0 WILL BE PRINTED!
[14,1] element, prob = 0.765625    (815,163)-(863,222) batch id : 0 WILL BE PRINTED!
[15,1] element, prob = 0.463379    (1025,185)-(1066,239) batch id : 0
[16,1] element, prob = 0.178223    (1272,240)-(1317,306) batch id : 0
[17,1] element, prob = 0.0708008    (663,182)-(706,238) batch id : 0
[18,1] element, prob = 0.0664062    (1073,186)-(1115,238) batch id : 0
[19,1] element, prob = 0.0395508    (879,176)-(921,228) batch id : 0
[20,1] element, prob = 0.0366211    (603,831)-(682,891) batch id : 0
[21,1] element, prob = 0.0283203    (1288,231)-(1335,313) batch id : 0
[22,1] element, prob = 0.0273438    (1302,606)-(1336,694) batch id : 0
[23,1] element, prob = 0.0239258    (853,172)-(897,225) batch id : 0
[24,1] element, prob = 0.0239258    (598,818)-(647,865) batch id : 0
[25,1] element, prob = 0.0229492    (535,788)-(577,847) batch id : 0
[26,1] element, prob = 0.0229492    (645,833)-(713,894) batch id : 0
[27,1] element, prob = 0.0219727    (856,155)-(895,204) batch id : 0
[28,1] element, prob = 0.0219727    (1016,200)-(1060,259) batch id : 0
[29,1] element, prob = 0.0209961    (182,102)-(228,159) batch id : 0
[30,1] element, prob = 0.0209961    (568,850)-(611,893) batch id : 0
[31,1] element, prob = 0.0209961    (1290,261)-(1341,351) batch id : 0
[32,1] element, prob = 0.0200195    (1010,146)-(1051,198) batch id : 0
[33,1] element, prob = 0.0200195    (630,181)-(670,231) batch id : 0
[34,1] element, prob = 0.0200195    (1285,633)-(1345,724) batch id : 0
[35,1] element, prob = 0.0200195    (560,827)-(625,904) batch id : 0
[36,1] element, prob = 0.0180664    (880,157)-(919,203) batch id : 0
[37,1] element, prob = 0.0180664    (79,547)-(124,601) batch id : 0
[38,1] element, prob = 0.0180664    (559,771)-(601,825) batch id : 0
[39,1] element, prob = 0.0180664    (572,766)-(648,830) batch id : 0
[40,1] element, prob = 0.0180664    (530,821)-(588,904) batch id : 0
[41,1] element, prob = 0.0170898    (598,846)-(651,892) batch id : 0
[42,1] element, prob = 0.0170898    (-21,815)-(62,914) batch id : 0
[43,1] element, prob = 0.0161133    (512,794)-(555,853) batch id : 0
[44,1] element, prob = 0.0161133    (272,852)-(319,893) batch id : 0
[45,1] element, prob = 0.0161133    (539,849)-(578,893) batch id : 0
[46,1] element, prob = 0.0161133    (1286,-1)-(1343,63) batch id : 0
[47,1] element, prob = 0.0161133    (517,765)-(595,862) batch id : 0
[48,1] element, prob = 0.0161133    (409,773)-(542,908) batch id : 0
[49,1] element, prob = 0.0161133    (375,23)-(1246,826) batch id : 0
[50,1] element, prob = 0.015625    (983,145)-(1021,188) batch id : 0
[51,1] element, prob = 0.015625    (1303,648)-(1337,715) batch id : 0
[52,1] element, prob = 0.015625    (629,848)-(671,894) batch id : 0
[53,1] element, prob = 0.015625    (38,41)-(663,852) batch id : 0
[54,1] element, prob = 0.0146484    (177,135)-(225,199) batch id : 0
[55,1] element, prob = 0.0146484    (1171,212)-(1214,271) batch id : 0
[56,1] element, prob = 0.0146484    (563,801)-(611,859) batch id : 0
[57,1] element, prob = 0.0146484    (658,818)-(706,872) batch id : 0
[58,1] element, prob = 0.0146484    (312,854)-(354,894) batch id : 0
[59,1] element, prob = 0.0146484    (1277,547)-(1344,704) batch id : 0
[60,1] element, prob = 0.0146484    (457,788)-(619,899) batch id : 0
[61,1] element, prob = 0.0146484    (1212,784)-(1360,908) batch id : 0
[62,1] element, prob = 0.0146484    (-47,632)-(209,940) batch id : 0
[63,1] element, prob = 0.0146484    (39,648)-(349,934) batch id : 0
[64,1] element, prob = 0.0146484    (1247,536)-(1426,1057) batch id : 0
[65,1] element, prob = 0.0136719    (1304,591)-(1337,656) batch id : 0
[66,1] element, prob = 0.0136719    (628,818)-(674,866) batch id : 0
[67,1] element, prob = 0.0136719    (350,850)-(393,893) batch id : 0
[68,1] element, prob = 0.0136719    (501,850)-(540,893) batch id : 0
[69,1] element, prob = 0.0136719    (138,102)-(209,182) batch id : 0
[70,1] element, prob = 0.0136719    (1288,316)-(1341,394) batch id : 0
[71,1] element, prob = 0.0136719    (495,823)-(553,900) batch id : 0
[72,1] element, prob = 0.0136719    (685,834)-(755,892) batch id : 0
[73,1] element, prob = 0.0136719    (-13,628)-(92,793) batch id : 0
[74,1] element, prob = 0.0136719    (-3,667)-(79,878) batch id : 0
[75,1] element, prob = 0.0136719    (323,790)-(504,902) batch id : 0
[76,1] element, prob = 0.0136719    (646,804)-(846,894) batch id : 0
[77,1] element, prob = 0.0136719    (1164,589)-(1509,984) batch id : 0
[78,1] element, prob = 0.0126953    (1,0)-(37,45) batch id : 0
[79,1] element, prob = 0.0126953    (276,138)-(320,198) batch id : 0
[80,1] element, prob = 0.0126953    (887,211)-(929,269) batch id : 0
[81,1] element, prob = 0.0126953    (1045,206)-(1087,257) batch id : 0
[ INFO ] Image out_0.bmp created!
[ INFO ] Execution successful

[ INFO ] This sample is an API example, for any performance measurements please use the dedicated benchmark_app tool
 

I attached the image file.

I presume that the age-gender-recognition command would be:

$ ~/build/armv7l/Release/object_detection_sample_ssd -m ~/build/face-detection-adas-0001.xml -d MYRIAD -i ~/Pictures/_DSC0523.jpg -m ~/build/age-gender-recognition-retail-0013.xml
 

But I get this error.

[ INFO ] InferenceEngine:
        API version ............ 2.1
        Build .................. custom_releases/2019/R3_cb6cad9663aea3d282e0e8b3e0bf359df665d5d0
        Description ....... API
Parsing input parameters
[ INFO ] Files were added: 1
[ INFO ]     /home/pi/Pictures/_DSC0523.jpg
[ INFO ] Loading Inference Engine
[ INFO ] Device info:
        MYRIAD
        myriadPlugin version ......... 2.1
        Build ........... 30677
[ INFO ] Loading network files:
        /home/pi/build/age-gender-recognition-retail-0013.xml
        /home/pi/build/age-gender-recognition-retail-0013.bin
[ INFO ] Preparing input blobs
[ INFO ] Batch size is 1
[ INFO ] Preparing output blobs
[ ERROR ] Can't find a DetectionOutput layer in the topology
 

I am running on a raspberrypi 3B using Rasparian OS. 

Thanks for your help.

Russell

[David_C_Intel]

Hi Russell,

Thank you for posting your questions. 

It seems you are running the wrong sample with the wrong models, as the object_detection_sample_ssd uses only ssd based models.

You can check how to run the age-gender-recognition model with the Interactive Face Detection C++ Demo here.

If you have additional questions, feel free to ask.

Best regards,

David