HI
I want to run the model which I have already trained by tensoeflow in my computer.
First, I downloaded the github : GitHub - jkjung-avt/tf_trt_models: TensorFlow object detection models accelerated with NVIDIA TensorRT (TF-TRT),
follow the steps and run “python3 camera_tf_trt.py --usb --model ssd_mobilenet_v1_coco –build” , it works.
Second, The model in tensorflow is trained by fast_rcnn_inception_v2 ,class is 5.
According to this website(TensorRT之TX2部署TensorFlow目标检测模型_宗而研之的博客-CSDN博客_tx2目标检测)
I put model.ckpt.data , model.ckpt.meta , model.ckpt.index into the file Github/tf_trt_model/data/fast_rcnn_inception_v2_egohands.
Board_name.pbtxt is my label file address is Github/tf_trt_model/data/board_name.pbtxt.
Open fast_rcnn_inception_v2_egohands.config and change class into 5.
finall,I run “python3 camera_tf_trt.py --usb --model faster_rcnn_inception_v2_egohands --labelmap data/board_name.pbtxt --num-classes 5 --build” but got an core_dump error.
convert_graph.cc:952] Engine my_trt_op_0 creation for segment 0, composed of 443 nodes succeeded.
Segmentation fault (core dumped)