few quotations on gpu version

[yaira24]

First of all good job on MTCNN GPU version!

why the image size need to configure up front in the constructor? is there a resin why you comment the mtcnn::~mtcnn()?

[PKUZHOU]

Q1: TensorRT doesn't support dynamic input size, so the image size should be configured previously. Q2: There is something wrong with the destructor

[yaira24]

thanks for the answer about the destructor, I change the code there and I don't have any issue mtcnn::~mtcnn(){

delete[] pnet_engine;
delete rnet_engine;
delete onet_engine;
for (int i = 0;i<scales_.size();i++)
{
    delete(simpleFace_[i]);
}
delete[] simpleFace_;
simpleFace_ = NULL;
delete refineNet;
delete outNet;
free(boxes_data);
CHECK(cudaFree(gpu_boxes_data));

}