Multicamera person reidentification using deepstream-mdx-perception-app

[edit_or]

Please provide complete information as applicable to your setup.

**• Hardware Platform (Jetson / GPU)**RTX4080
• DeepStream Version7.1
I am exploring how to implement person reidentification in multiple cameras.
I am aware of Multi-Camera Tracking for retail store. That is quite heavy to implement.
I am looking for light weight version.
So I found this deepstream-mdx-perception-app. The application detect people and sgie is extract features. My query is i want to use that features for reidentification when normal tracking with NvDCF lost tracking a person.
My queries are
(1)can i use the app deepstream-mdx-perception-app for multicamera people tracking
(2)how long I can keep the features so that when new id is assigned, I will use that feature to reidentify?

[kesong]

Yes, you can use ReID vector for reidentify. It depends on your use case. If your use case need long time reidentify, you need keep the features longer.

[edit_or]

Can use for multiple cctv tracking?

[kesong]

If you don’t want to use MTMC as you mentioned, you need develop multiple camera tracking application by yourself.

[yingliu]

**There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks**