0

I'm parsing MobileNet-SSD caffe Model from https://github.com/chuanqi305/MobileNet-SSD using TensorRT caffe parser. I use Jetpack 3.3 in Jetson TX2 as the platform. How to fix this error from nvinfer library? There is even a TensorRT class nvinfer1::plugin::PriorBoxParameters in the API. But this layer handle seems to be not defined. Is there any error in the layer param names and format? Or do we have to create a custom layer plugin for this one. Does someone have a implementation for this?

The error is as follows,

[libprotobuf ERROR google/protobuf/text_format.cc:298] Error parsing text-format ditcaffe.NetParameter: 1245:18: Message type "ditcaffe.LayerParameter" has no field named "prior_box_param".
ERROR: CaffeParser: Could not parse deploy file

The source code line relevant for this error,

m_network = m_builder->createNetwork();
m_parser = createCaffeParser();
const IBlobNameToTensor* blobNameToTensor = m_parser->parse(deployFpath, modelFpath, *network, DataType::kFLOAT);

Error given layer in prototxt file,

layer {
  name: "conv11_mbox_priorbox"
  type: "PriorBox"
  bottom: "conv11"
  bottom: "data"
  top: "conv11_mbox_priorbox"
  prior_box_param {
    min_size: 60.0
    aspect_ratio: 2.0
    flip: true
    clip: false
    variance: 0.1
    variance: 0.1
    variance: 0.2
    variance: 0.2
    offset: 0.5
  }
}
Chinthaka Gamanayake
  • 11
  • 3

1 Answers1

0

You have to remove all the blocks that have _param from the ptototxt file, then you should removed layers using the plugin API. Visit the discussion on this link for clear understanding: https://github.com/chenzhi1992/TensorRT-SSD/issues/5

Mukesh Karanth
  • 11
  • 2