Real time Object Detection - Python 2 vs Python 3?

Robotics & Edge Computing Jetson & Embedded Systems Jetson AGX Xavier
1

Using this to install open CV and Tensorflow - https://github.com/naisy/JetsonXavier

And this app to run the object detection - https://github.com/naisy/realtime_object_detection

it works great, but only works in Python 2, is there any simple way to get it to work in Python 3?

Thanks!

2

Hi,

We have built TensorFlow for both python-2 and python-3.
Please check this topic for the package link:
[url]https://devtalk.nvidia.com/default/topic/1041712/jetson-agx-xavier/tensorflow-wheel-for-jetpack-4-0-/[/url]

Thanks.

3

We also have a repository on GitHub that demonstrates accelerated use of models from the TensorFlow Object Detection API.

[url]https://github.com/NVIDIA-Jetson/tf_trt_models[/url]

This includes a setup script that simplifies installation for Python 2 and 3. This repository currently covers a subset of the SSD models from the TensorFlow object detection API. We have tested that the following models work using Python 2 and 3 with some modifications applied.

ssd_mobilenet_v1_coco
ssd_mobilenet_v2_coco
ssd_inception_v2_coco
ssd_resnet_50_fpn_coco

Please let me know if you’re interested in using one of the above models and I’m happy to provide more details.

4

Thanks, everyone, it’s going to be a few days until I can get back to this portion of the project so I’ll let you know then. Thanks again!