[robotics-worldwide] [software] Bonnetal: Deep Learning Framework for various Image Recognition Tasks. Photogrammetry and Robotics Lab, University of Bonn. Andres Milioto and Cyrill Stachniss

Andres Milioto amilioto at uni-bonn.de
Mon Jul 1 06:37:53 PDT 2019


Dear Colleagues,

In early 2018 we released Bonnet, which is a real-time, robotics oriented
semantic segmentation framework using Convolutional Neural Networks (CNNs).
Bonnet provides an easy pipeline to add architectures and datasets for
semantic segmentation, in order to train and deploy CNNs on a robot. Bonnet
contains a full training pipeline in Python using Tensorflow and OpenCV,
and it also some C++ apps to deploy a CNN in ROS and standalone. The C++
library is made in a way which allows adding other backends (such as
TensorRT) for fast inference. The Bonnet whitepaper was published this year
at ICRA 2019 in Montreal.

Back then, most of my research was in the field of semantic segmentation,
so that was what the framework was therefore tailored specifically to do.
Since then, we have found a way to make things more general, allowing for a
suite of other tasks, like classification, detection, instance and semantic
segmentation, feature extraction, counting, etc. Hence, the new name of
this new framework: "Bonnetal", reflects that this is nothing but the old
Bonnet, and then some. Hopefully, the explict "et.al." will also spawn
further collaboration and many pull requests! The nice thing about Bonnetal
(as with Bonnet) is that all architectures can be used with our C++
library, which also has a ROS wrapper so that you don't have to code at
all. And all of the supported architectures are tested using NVIDIA
TensorRT so that you get that extra juice out of your Jetson or GPU (INT8
quantization is also included).

The network shown in the following video is a MobilenetsV2 with a small
Atrous Spatial Pyramid Pooling module attached, running quantized to INT8
for fast inference, at VGA resolution (200fps):
https://urldefense.proofpoint.com/v2/url?u=https-3A__www.youtube.com_watch-3Fv-3Dndp0L2oPlIE&d=DwIBaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=a5mn2Mo-_GHffnbVYOnJdi6xKlN4Unn1LcMRcY4psZE&s=5m7_sLoBURDOIyfWLXdiWT3tadLBoxBu1E2W-AM5ORY&e= 

You can access the code here: https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_PRBonn_bonnetal&d=DwIBaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=0w3solp5fswiyWF2RL6rSs8MCeFamFEPafDTOhgTfYI&m=a5mn2Mo-_GHffnbVYOnJdi6xKlN4Unn1LcMRcY4psZE&s=37tI_oltBY5RBRBM5SVcSmSOi5GmpCfewC-augOqIkY&e= 

Best regards,

Andres Milioto
Research Assistant and Doctoral Student
IGG - Photogrammetry
University of Bonn
Nussallee 15
53115 Bonn
Germany


More information about the robotics-worldwide mailing list