openvinotoolkit/open_model_zoo

[Suggestion] Improve Pedestrain Trackor C++ Demo

Open

#2,260 opened on Mar 31, 2021

View on GitHub
 (6 comments) (1 reaction) (0 assignees)Python (1,334 forks)batch import
demofeaturehelp wanted

Repository metrics

Stars
 (3,759 stars)
PR merge metrics
 (No merged PRs in 30d)

Description

Hi, OpenVINO guys, I learn a lot from your Pedestrain Trackor demo, however I wonder why some intel model can't be supported in this demo. Such as person-detection-0203, person-detection-0106. Only person-detection-0201/0202 are supported.

I notice it's because the detection output is a little different, could you please improve the demo to let's try more detection models, because I want to try cascade-RCNN model to get better accuracy.

Thanks.

Contributor guide

Research direction
Examine the pedestrian tracker C++ demo source code in the open model zoo repository, focusing on how it currently parses detection outputs for the supported models (person detection 0201/0202). Compare the output format of unsupported models like person detection 0203 and person detection 0106, which may use different output layers or post processing. Modify the demo to dynamically handle multiple output formats, possibly by adding a configuration parameter or by detecting the model type. Refer to the OpenVINO model API documentation for output layer details.
Tech stack
cpppython
Domain
backenddesktop
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Mostly clear
Prerequisites
C++OpenVINOmodel inference basics
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
40