# Deepstream sample to usefor our solution

**URL:** <https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030>\
**Category:** Jetson Nano\
**Created:** [January 16, 2021, 9:47am UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030 "2021-01-16T09:47:49Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![bvgohmslf](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/bvgohmslf/32/31953_2.png) [@bvgohmslf](https://forums.developer.nvidia.com/u/bvgohmslf)\
**Post date:** [January 16, 2021, 9:47am UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/1 "2021-01-16T09:47:49Z")

</div>

Hi,

We currently have lags when using opencv dlib on jetson nano, hence we are making a switch to deepstream.

We need suggestions on an appropriate deepstream sample app working with taking a video stream from a remote camera to jetson nano process face recognition and people detection while taking picture images and short video recordings to the cloud for mobile notifications.

A python pipeline would be a huge advantage for us.

Thanks

---

<div class="post-metadata">

**Author:** ![shubham.shah09](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@shubham.shah09](https://forums.developer.nvidia.com/u/shubham.shah09)\
**Post date:** [January 16, 2021, 11:40am UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/2 "2021-01-16T11:40:36Z")

</div>

only face detection or recognition as well?

---

<div class="post-metadata">

**Author:** ![bvgohmslf](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/bvgohmslf/32/31953_2.png) [@bvgohmslf](https://forums.developer.nvidia.com/u/bvgohmslf)\
**Post date:** [January 17, 2021, 12:04am UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/3 "2021-01-17T00:04:50Z")

</div>

Recognition to be precise such as recognizing visitors and so on.

---

<div class="post-metadata">

**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\
**Post date:** [January 18, 2021, 3:18am UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/5 "2021-01-18T03:18:02Z")

</div>

Hi,

Here is our sample for python:

> **[GitHub - NVIDIA-AI-IOT/deepstream\_python\_apps: DeepStream SDK Python bindings and sample...](https://github.com/NVIDIA-AI-IOT/deepstream_python_apps)**
>
> DeepStream SDK Python bindings and sample applications

You can check _deepstream-test2_ for a detection+classification(recognition) pipeline.  
Thanks.

---

<div class="post-metadata">

**Author:** ![bvgohmslf](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/bvgohmslf/32/31953_2.png) [@bvgohmslf](https://forums.developer.nvidia.com/u/bvgohmslf)\
**Post date:** [January 18, 2021, 12:50pm UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/6 "2021-01-18T12:50:19Z")

</div>

How can I focus on just people detection without including the vehicle classification and so on? I understand I will need to edit the config file but don’t want to break anything so I ask this.

Thanks

---

<div class="post-metadata">

**Author:** ![AastaLLL](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/aastalll/32/14043_2.png) [@AastaLLL](https://forums.developer.nvidia.com/u/AastaLLL)\
**Post date:** [January 20, 2021, 5:33am UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/7 "2021-01-20T05:33:20Z")

</div>

Hi,

You will need to replace the primary GIE with a face detection model.  
And the secondary GIE with a face recognition(classification) model.

Do you already have a detection and classification model for your use case?  
If not, you can train one with our toolkit below:

> **[NVIDIA TAO Toolkit](https://developer.nvidia.com/tao-toolkit)**
>
> Leverage the power of transfer learning to fine-tune NVIDIA pretrained models.

Thanks.

---

<div class="post-metadata">

**Author:** ![nadeemm](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/nadeemm/32/367020_2.png) [@nadeemm](https://forums.developer.nvidia.com/u/nadeemm)\
**Post date:** [October 14, 2021, 8:49pm UTC](https://forums.developer.nvidia.com/t/deepstream-sample-to-usefor-our-solution/166030/9 "2021-10-14T20:49:56Z")

</div>

Closed on October 14, 2021, 8:49pm UTC