I add a deconvolution layer before one input of roiproposal plugin in samplefastercnn.
layer {
name: “convf_up2”
type: “Deconvolution”
bottom: “convf”
top: “convf_up2”
param {
lr_mult: 0.0
decay_mult: 0.0
}
convolution_param {
num_output: 256
bias_term: false
pad: 1
kernel_size: 4
group: 256
stride: 2
weight_filler {
type: “bilinear”
}
but I got one error :RPROIPluginsampleFasterRCNN: NvPluginFasterRCNN.cu:83: virtual void nvinfer1::plugin::RPROIPlugin::configure(const nvinfer1::Dims*, int, const nvinfer1::Dims*, int, int): Assertion `inputDims[0].d[1] == inputDims[1].d[1] && inputDims[0].d[1] == inputDims[2].d[1]’ failed.
[1] 7214 abort (core dumped) ./bazel-bin/sampleFasterRCNN/sampleFasterRCNN/sampleFasterRCNN
have any one met this problem before ? could Nvidia open source the implementation of proposalroi plugin?