Tx1 DetectNet failure on "Python" layer

the DetectNet on Jetson Tx1 would report "Python"layer unknown when I have the bellow layer in the deploy.prototxt:

layer {
name: “cluster”
type: “Python”
bottom: “coverage”
bottom: “bboxes”
top: “bbox-list”
python_param {
module: “caffe.layers.detectnet.clustering”
layer: “ClusterDetections”
param_str: “1248, 352, 16, 0.6, 3, 0.02, 22, 1”
}
}

How can I fix this?

Hi,

Thanks for your question.

Python layer is a special layer only implemented in nvcaffe.
Please remove this layer and re-implement it with c++ or opencv on TX1.

Example can be found in GitHub - dusty-nv/jetson-inference: Hello AI World guide to deploying deep-learning inference networks and deep vision primitives with TensorRT and NVIDIA Jetson.
Find more information in this topic: https://devtalk.nvidia.com/default/topic/998570

Thanks.