# Deepstream\_pose\_estimation doesn't show pose estimation skeleton

**URL:** <https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058>\
**Category:** DeepStream SDK\
**Created:** [February 22, 2021, 2:35pm UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058 "2021-02-22T14:35:37Z")\
**Posts on this page:** 10\
**Page:** 1

<div class="post-metadata">

**Author:** ![edit\_or](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@edit\_or](https://forums.developer.nvidia.com/u/edit_or)\
**Post date:** [February 22, 2021, 2:35pm UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/1 "2021-02-22T14:35:37Z")

</div>

I am testing [deepstream\_pose\_estimation](https://github.com/NVIDIA-AI-IOT/deepstream_pose_estimation).  
With provided [onnx file](https://github.com/NVIDIA-AI-IOT/deepstream_pose_estimation/blob/master/pose_estimation.onnx), I see some detection.

But when I use the [following two models](https://github.com/NVIDIA-AI-IOT/trt_pose), detection just show small white circles only, no green lines.

```
resnet18_baseline_att_224x224_A
densenet121_baseline_att_256x256_B

```

What should I change?  
I managed to work with mp4 file. The original program uses h264 raw stream.  
Can detection accuracy drop using mp4 format?

---

<div class="post-metadata">

**Author:** ![bcao](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/bcao/32/14043_2.png) [@bcao](https://forums.developer.nvidia.com/u/bcao)\
**Post date:** [February 24, 2021, 1:17am UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/3 "2021-02-24T01:17:11Z")

</div>

Hey could you share your setup with us?

---

<div class="post-metadata">

**Author:** ![edit\_or](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@edit\_or](https://forums.developer.nvidia.com/u/edit_or)\
**Post date:** [February 24, 2021, 4:11am UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/4 "2021-02-24T04:11:10Z")

</div>

Which set up file? Is it config file?

---

<div class="post-metadata">

**Author:** ![edit\_or](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@edit\_or](https://forums.developer.nvidia.com/u/edit_or)\
**Post date:** [February 26, 2021, 1:09am UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/5 "2021-02-26T01:09:57Z")

</div>

I sent by private message all files (cpp file that can run mp4 file, models, config file and test video).  
Only pose\_estimation.onnx works (still not very good). The other models don’t work.

---

<div class="post-metadata">

**Author:** ![edit\_or](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@edit\_or](https://forums.developer.nvidia.com/u/edit_or)\
**Post date:** [February 26, 2021, 2:58pm UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/6 "2021-02-26T14:58:37Z")

</div>

Do you receive my files? Thanks

---

<div class="post-metadata">

**Author:** ![edit\_or](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@edit\_or](https://forums.developer.nvidia.com/u/edit_or)\
**Post date:** [March 1, 2021, 2:38am UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/7 "2021-03-01T02:38:08Z")

</div>

I have sent my files in message. Have u received?

---

<div class="post-metadata">

**Author:** ![bcao](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/bcao/32/14043_2.png) [@bcao](https://forums.developer.nvidia.com/u/bcao)\
**Post date:** [March 1, 2021, 1:56pm UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/8 "2021-03-01T13:56:02Z")

</div>

> [@edit\_or](#):
>
> Which set up file? Is it config file?

the device/DS version.

---

<div class="post-metadata">

**Author:** ![edit\_or](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@edit\_or](https://forums.developer.nvidia.com/u/edit_or)\
**Post date:** [March 2, 2021, 1:19am UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/9 "2021-03-02T01:19:21Z")

</div>

I am using Xavier. Jetpack is 4.4 and TensorRT 7 and Deepstream 5.0

Thanks

---

<div class="post-metadata">

**Author:** ![bcao](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/bcao/32/14043_2.png) [@bcao](https://forums.developer.nvidia.com/u/bcao)\
**Post date:** [March 16, 2021, 7:11am UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/10 "2021-03-16T07:11:18Z")

</div>

I think 1st, you need to make sure if the model itself can give a better output.  
2nd, if the model’s output is well, then you need to check if the input w/o DS pipeline is the same, you can dump the model’s input via [DeepStream SDK FAQ - #9 by mchi](https://forums.developer.nvidia.com/t/deepstream-sdk-faq/80236/9)

---

<div class="post-metadata">

**Author:** ![TomNVIDIA](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/tomnvidia/32/14181_2.png) [@TomNVIDIA](https://forums.developer.nvidia.com/u/TomNVIDIA)\
**Post date:** [October 12, 2021, 4:14pm UTC](https://forums.developer.nvidia.com/t/deepstream-pose-estimation-doesnt-show-pose-estimation-skeleton/169058/12 "2021-10-12T16:14:52Z")

</div>

Closed on October 12, 2021, 4:14pm UTC


