Skip to content
No description, website, or topics provided.
Python Dockerfile Shell
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
packages fix Dec 29, 2019
utils init Dec 29, 2019
.gitignore
Dockerfile
LICENSE fix Dec 29, 2019
README.md
build.sh fix Dec 29, 2019
dogs.jpg add image Dec 29, 2019
trt_ssd_benchmark.py

README.md

How to run SSD Mobilenet V2 object detection on Jetson Nano at 20+ FPS | DLology

First, make sure you have flashed the latest JetPack 4.3 on your Jetson Nano development SD card.

Run the docker

docker run --runtime nvidia --network host --privileged -it docker.io/zcw607/trt_ssd_r32.3.1:0.1.0

Then run this command to benchmark the inference speed.

python3 trt_ssd_benchmark.py

To build the docker yourself

To build the docker on Jetson nano or any other arm64 board.

docker build -t docker.io/<your user name>/trt_ssd_r32.3.1:0.1.0 . -f ./Dockerfile
docker push docker.io/<your user name>/trt_ssd_r32.3.1:0.1.0

To build the docker on a x86 machine, requires podman and qemu installed.

podman build -v /usr/bin/qemu-aarch64-static:/usr/bin/qemu-aarch64-static -t docker.io/<your user name>/trt_ssd_r32.3.1:0.1.0 . -f ./Dockerfile
podman push docker.io/<your user name>/trt_ssd_r32.3.1:0.1.0
You can’t perform that action at this time.