Git Product home page Git Product logo

blazepose_openvino's Introduction

blazepose_openvino

BlazePose model test program for Intel Distribution of Intel OpenVINO toolkit

WIP

The BlazePose OpenVINO IR model can be obtained from PINTO model zoo.

Example (Upper body model)

Example

Example (Full body model)

Example-full

Configuration / Options

Current program doesn't support command line options.
You can modify the source code to change input media/file and device to use for inferencing. Those configurations can be found on top of the main.cpp code.

#define FULL_POSE
#include "blazepose.h"

#define IMAGE_INPUT    (1)
#define VIDEO_INPUT    (2)
#define CAM_INPUT      (3)

#ifndef FULL_POSE
const std::string MODEL_POSE_DET = "../pose_detection/128x128/FP32/pose_detection";
const std::string MODEL_LM_DET   = "../pose_landmark_upper_body/256x256/FP32/pose_landmark_upper_body";
#else
const std::string MODEL_POSE_DET = "../pose_detection_full/128x128/FP16/pose_detection_full";
const std::string MODEL_LM_DET   = "../pose_landmark_full_body/256x256/FP16/pose_landmark_full_body";
#endif

// INPUT_TYPE = { IMAGE_INPUT | VIDEO_INPUT | CAM_INPUT }
#define INPUT_TYPE    VIDEO_INPUT
const std::string INPUT_FILE = "../boy.mp4";                    /* Image or movie file */

// 'output.mp4' will be generated when this macro is defined and the input source is either one of VIDEO_INPUT or CAM_INPUT
#define VIDEO_OUTPUT
#define VIDEO_SIZE    (400)             /* output video size = (VIDEO_SIZE, VIDEO_SIZE) */

// Device to use for inferencing. Possible options = "CPU", "GPU", "MYRIAD", "HDDL", "HETERO:FPGA,CPU", ...
const std::string DEVICE_PD = "CPU";
const std::string DEVICE_LM = "CPU";

// ** Define or Undefine to control the items to display
//#define RENDER_ROI
#define RENDER_TIME
#define RENDER_POINTS

blazepose_openvino's People

Contributors

yas-sim avatar yassim-intel avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar

blazepose_openvino's Issues

Missing models.

Code needs following files of models:
pose_detection/128x128/FP32/pose_detection.xml
pose_detection/128x128/FP32/pose_detection.bin
pose_detection_full/128x128/FP16/pose_detection_full.xml
pose_detection_full/128x128/FP16/pose_detection_full.bin
pose_landmark_upper_body/256x256/FP32/pose_landmark_upper_body.xml
pose_landmark_upper_body/256x256/FP32/pose_landmark_upper_body.bin

You write in README:

The BlazePose OpenVINO IR model can be obtained from PINTO model zoo.

Could you be so kind to specify exactly which models from PINTO model zoo are meant and how to obtain these models in format needed?

tensor scores layers "StatefulPartitionedCall/functional_1/classificators/concat " crash at setPrecision(ie::Precision::FP32) without any outputed error.

Hello, thank you for sharing this. It will certainly helped me to quick dive into blazepose.

Im moreover stuck at a crash at the line 570 of the main file.
output_info->setPrecision(ie::Precision::FP32);
Program exit without returning any error.
I understand that input models must have been compiled with FP32 precision.
I loaded them from PINTO0 Model Zoo and defined inputs as :
MODEL_POSE_DET = .\053_Blazepose\03_pose_landmark_full_body\saved_model_openvino\openvino\FP32\saved_model
MODEL_LM_DET = .\053_Blazepose\01_pose_detection\saved_model_openvino\openvino\FP32\saved_model
They seem to respect precision criteria.

Could you give me a hinch about what Im doing wrong?

Regards,

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.