myih

@linux-devil @Maxfashko Any suggestion on how to compile this code? The CMakeList is incomplete. Thank you.

@OMG59E @chenzhi1992 I was able to get it running with TensorRT 4.0 but the results are incorrect. Around 5.6ms on 1080Ti without implementing depthwise conv plugin (use group conv). If...

@xmglin I think I met this error before, you need to match the IPlugin layers' name in the .prototxt with pluginimplemenation.cpp/h's layer, this repo only provides VGG-SSD's pluginimplemenation so you...

@OMG59E I got the wrong inference time as I included the overheads... Using the timeInference included in tensorNet: Time over total layers: 2.561 conv0 + conv0/relu 0.090ms conv1/dw + conv1/dw/relu...

@paghdv @chenzhi1992 Did you solve this problem? I have the same error `Assertion 'numPriors*numLocClasses*4 == inputDims[param.inputOrder[0]].d[0]' failed.` I've already ran TensorRT-SSD successfully and now trying to run Mobilenet-SSD with TensorRT3....

@paghdv Thanks for the fast reply. I'm able to use the createSSDDetectionOutputPlugin to run VGG-SSD so maybe it's not the detection-layer's problem? What version of TensorRT and cudnn are you...

@paghdv I solve the error by updating to TensorRT4 and specify inputOrder = {0, 1, 2}, but the detection output result is incorrect. Were you able to get correct output...

@paghdv Thanks for the reply. @zbw4034 What model and weights are you trying to run? I solve this problem after setting _inputOrder = {0, 1, 2}_ in createSSDDetectionOutputPlugin, also you...

@paghdv I use this implementation [Teoge/tensorrt-ssd-easy](https://github.com/Teoge/tensorrt-ssd-easy ) , it works for vgg-ssd and I checked the cuda code seems legit.

Damn I totally forgot about Weiliu and chuanqi305 use different image loader... Now it can run on 1080ti with ~150fps (without depthwise convolution implemented, is it normal?)