topic Hi Ranganadh, in Intel® Distribution of OpenVINO™ Toolkit

NCS-2 in Intel OpenVINO Interactive-face-detection-demo

Ranganadh_K_Intel — Mon, 28 Oct 2019 10:12:09 GMT

Hi,

I am trying to run the C++ Interactive-face-detection-demo in Ubuntu 16.04 in Intel dev-cloud using the cpp reference samples.

In the cpp interactive-face-detection reference sample they are running by providing all models at a time using below command.

./face_detector -i $INPUT_FILE -no_wait \
-m models/Transportation/object_detection/face/pruned_mobilenet_reduced_ssd_shared_weights/dldt/face-detection-adas-0001${FPEXT}.xml \
-m_hp models/Transportation/object_attributes/headpose/vanilla_cnn/dldt/head-pose-estimation-adas-0001${FPEXT}.xml \
-m_ag models/Retail/object_attributes/age_gender/dldt/age-gender-recognition-retail-0013${FPEXT}.xml \
-m_em models/Retail/object_attributes/emotions_recognition/0003/dldt/emotions-recognition-retail-0003${FPEXT}.xml \
-d $DEVICE -d_hp $DEVICE -d_ag $DEVICE -d_em $DEVICE \
-o $OUTPUT_FILE

It is working fine. But If i tried to run the above sample with only face-detection model with the below command.

./face_detector -i $INPUT_FILE -no_wait \ -m models/Transportation/object_detection/face/pruned_mobilenet_reduced_ssd_shared_weights/dldt/face-detection-adas-0001-fp16.xml -d MYRIAD

I am getting the below error ..

[ ERROR ] Cannot make shared blob! The blob type cannot be used to store objects of current precision
/opt/intel/openvino_2019.1.094/deployment_tools/inference_engine/include/details/ie_exception_conversion.hpp:71

Appreciated for any reply on this issue.

Hi Ranganadh,

JesusE_Intel — Mon, 28 Oct 2019 21:30:48 GMT

Hi Ranganadh,

I tested the interactive face detection demo just using the face detection model and it work fine with the following command on OpenVINO 2019 R3.1. I'm not sure if you can upgrade your version on the Intel DevCloud, however, you should be able find the answer on the Intel DevCloud support forum.

./interactive_face_detection_demo -i cam -m /opt/intel/openvino/deployment_tools/open_model_zoo/tools/downloader/intel/face-detection-adas-0001/FP16/face-detection-adas-0001.xml -d MYRIAD

Regards,

Jesus

Hi Jesus,

Ranganadh_K_Intel — Tue, 29 Oct 2019 03:59:16 GMT

Hi Jesus,

Thank you so much for your quick response.

I m trying to run with flag -o , for storing output in a file in disk, using the below command

./interactive_face_detection_demo -i ~/smart_video_validation/iot-devcloud/smart-video-workshop/advanced-video-analytics/cpp/faces-recognition-walking-and-pause.mp4 -m ~/smart_video_validation/iot-devcloud/smart-video-workshop/advanced-video-analytics/cpp/models/Transportation/object_detection/face/pruned_mobilenet_reduced_ssd_shared_weights/dldt/face-detection-adas-0001.xml -o ~/results/out.mp4

InferenceEngine:
API version ............ 1.6
Build .................. custom_releases/2019/R1_c9b66a26e4d65bb986bb740e73f58c6e9e84c7c2
[ INFO ] Parsing input parameters
[ INFO ] Reading input
OpenCV: FFMPEG: tag 0x56555949/'IYUV' is not supported with codec id 14 and format 'mp4 / MP4 (MPEG-4 Part 14)'
[mp4 @ 0x51957c0] Could not find tag for codec rawvideo in stream #0, codec not currently supported in container

(interactive_face_detection_demo:43262): GStreamer-CRITICAL **: gst_element_make_from_uri: assertion 'gst_uri_is_valid (uri)' failed.

Please find the attached screenshot for good understanding about issue.

Hi Ranganadh,

JesusE_Intel — Wed, 30 Oct 2019 17:57:08 GMT

Hi Ranganadh,

I was able to save the inference results to a video file (mp4) with the -o flag on the example without any issues on OpenVINO 2019 R3.1. The issue may be related to the OpenVINO 2019 R1 or the configuration of the Intel DevCloud. I recommend contacting the Intel DevCloud support for this issue.

https://colfaxresearch.com/discussion/forum/colfax-cluster/

Regards,

Jesus