Ardian Umam

In object detection, we need at least 3 tensors to be called: (i) input, (ii) class prediction and (iii) box prediction. In this case, we need to know those tensor...

You can, e.g., take a look on the code that calls the frozen model (*.pb file). There should be corresponding tensor output names needed to call. Or, you can ask...

Hi, Do you: (i) run all the block 2 code of [this code file](https://github.com/ardianumam/Tensorflow-TensorRT/blob/master/7_optimizing_YOLOv3_using_TensorRT.ipynb), or (ii) only run function `read_pb_graph("./model/YOLOv3/yolov3_gpu_nms.pb`? If (i), yes, it takes longer time since you also...

@fugjo16 : Do you convert the `frozen_model.pb` to `TRT_model.pb` in Desktop, then, you use it in Jetson TX2? Because I ever do the similar, and yes, it takes very long...

@fugjo16 : I just knew about that. I'll try later in my TX2 too, and update here soon. Thanks. Yes, probably that's the cause.

Hi @fugjo16 : I just tried in my TX2, and yes, it took about 15 minutes to only read the ``, meanwhile reading the native tensorflow model `.pb` needs only...

@filipski : thanks for the info. I'll give it a try.

> Why not at the end of model Because after the network output, we still have further operation, such as loss calculation, etc. > what `:0` is It is tensorflow...

Those only tensor names. Everyone can name the tensor any name he wants. Those output in object detector correspond to the outputs of classification and bonding box predictions.

> Hi, > I optimized my trained model (1 class), the ssdlite_mobilenetv2 and ssd_resnet50, with TensorRT, but the performance did't improve significantly, I reach from 0.12 sec to 0.11 sec...