<img src="https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/deep-vision-header.jpg" width="100%">
# Deploying Deep Learning
Welcome to our instructional guide for inference and realtime [DNN vision](#api-reference) library for NVIDIA **[Jetson Nano/TX1/TX2/Xavier NX/AGX Xavier](http://www.nvidia.com/object/embedded-systems.html)**.
This repo uses NVIDIA **[TensorRT](https://developer.nvidia.com/tensorrt)** for efficiently deploying neural networks onto the embedded Jetson platform, improving performance and power efficiency using graph optimizations, kernel fusion, and FP16/INT8 precision.
Vision primitives, such as [`imageNet`](c/imageNet.h) for image recognition, [`detectNet`](c/detectNet.h) for object detection, and [`segNet`](c/segNet.h) for semantic segmentation, inherit from the shared [`tensorNet`](c/tensorNet.h) object. Examples are provided for streaming from live camera feed and processing images. See the **[API Reference](#api-reference)** section for detailed reference documentation of the C++ and Python libraries.
<img src="https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/deep-vision-primitives.png" width="800">
Follow the [Hello AI World](#hello-ai-world) tutorial for running inference and transfer learning onboard your Jetson, including collecting your own datasets and training your own models. It covers image classification, object detection, and segmentation.
### Table of Contents
* [Hello AI World](#hello-ai-world)
* [Video Walkthroughs](#video-walkthroughs)
* [API Reference](#api-reference)
* [Code Examples](#code-examples)
* [Pre-Trained Models](#pre-trained-models)
* [System Requirements](#recommended-system-requirements)
* [Change Log](CHANGELOG.md)
> > [Jetson Nano 2GB](https://developer.nvidia.com/embedded/jetson-nano-2gb-developer-kit) and JetPack 4.4.1 is now supported in the repo. <br/>
> > Try the new [Re-training SSD-Mobilenet](docs/pytorch-ssd.md) object detection tutorial! <br/>
> > See the [Change Log](CHANGELOG.md) for the latest updates and new features. <br/>
## Hello AI World
Hello AI World can be run completely onboard your Jetson, including inferencing with TensorRT and transfer learning with PyTorch. The inference portion of Hello AI World - which includes coding your own image classification and object detection applications for Python or C++, and live camera demos - can be run on your Jetson in roughly two hours or less, while transfer learning is best left to leave running overnight.
#### System Setup
* [Setting up Jetson with JetPack](docs/jetpack-setup-2.md)
* [Running the Docker Container](docs/aux-docker.md)
* [Building the Project from Source](docs/building-repo-2.md)
#### Inference
* [Classifying Images with ImageNet](docs/imagenet-console-2.md)
* [Using the ImageNet Program on Jetson](docs/imagenet-console-2.md)
* [Coding Your Own Image Recognition Program (Python)](docs/imagenet-example-python-2.md)
* [Coding Your Own Image Recognition Program (C++)](docs/imagenet-example-2.md)
* [Running the Live Camera Recognition Demo](docs/imagenet-camera-2.md)
* [Locating Objects with DetectNet](docs/detectnet-console-2.md)
* [Detecting Objects from Images](docs/detectnet-console-2.md#detecting-objects-from-the-command-line)
* [Running the Live Camera Detection Demo](docs/detectnet-camera-2.md)
* [Coding Your Own Object Detection Program](docs/detectnet-example-2.md)
* [Semantic Segmentation with SegNet](docs/segnet-console-2.md)
* [Segmenting Images from the Command Line](docs/segnet-console-2.md#segmenting-images-from-the-command-line)
* [Running the Live Camera Segmentation Demo](docs/segnet-camera-2.md)
#### Training
* [Transfer Learning with PyTorch](docs/pytorch-transfer-learning.md)
* Classification/Recognition (ResNet-18)
* [Re-training on the Cat/Dog Dataset](docs/pytorch-cat-dog.md)
* [Re-training on the PlantCLEF Dataset](docs/pytorch-plants.md)
* [Collecting your own Classification Datasets](docs/pytorch-collect.md)
* Object Detection (SSD-Mobilenet)
* [Re-training SSD-Mobilenet](docs/pytorch-ssd.md)
* [Collecting your own Detection Datasets](docs/pytorch-collect-detection.md)
#### Appendix
* [Camera Streaming and Multimedia](docs/aux-streaming.md)
* [Image Manipulation with CUDA](docs/aux-image.md)
* [Deep Learning Nodes for ROS/ROS2](https://github.com/dusty-nv/ros_deep_learning)
## Video Walkthroughs
Below are screencasts of Hello AI World that were recorded for the [Jetson AI Certification](https://developer.nvidia.com/embedded/learn/jetson-ai-certification-programs) course:
| Description | Video |
|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| <a href="https://www.youtube.com/watch?v=QXIwdsyK7Rw&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=9" target="_blank">**Hello AI World Setup**</a><br/>Download and run the Hello AI World container on Jetson Nano, test your camera feed, and see how to stream it over the network via RTP. | <a href="https://www.youtube.com/watch?v=QXIwdsyK7Rw&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=9" target="_blank"><img src=https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/thumbnail_setup.jpg width="750"></a> |
| <a href="https://www.youtube.com/watch?v=QatH8iF0Efk&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=10" target="_blank">**Image Classification Inference**</a><br/>Code your own Python program for image classification using Jetson Nano and deep learning, then experiment with realtime classification on a live camera stream. | <a href="https://www.youtube.com/watch?v=QatH8iF0Efk&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=10" target="_blank"><img src=https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/thumbnail_imagenet.jpg width="750"></a> |
| <a href="https://www.youtube.com/watch?v=sN6aT9TpltU&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=11" target="_blank">**Training Image Classification Models**</a><br/>Learn how to train image classification models with PyTorch onboard Jetson Nano, and collect your own classification datasets to create custom models. | <a href="https://www.youtube.com/watch?v=sN6aT9TpltU&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=11" target="_blank"><img src=https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/thumbnail_imagenet_training.jpg width="750"></a> |
| <a href="https://www.youtube.com/watch?v=obt60r8ZeB0&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=12" target="_blank">**Object Detection Inference**</a><br/>Code your own Python program for object detection using Jetson Nano and deep learning, then experiment with realtime detection on a live camera stream. | <a href="https://www.youtube.com/watch?v=obt60r8ZeB0&list=PL5B692fm6--uQRRDTPsJDp4o0xbzkoyf8&index=12" target="_blank"><img src=https://github.com/dusty-nv/jetson-inference/raw/master/docs/images/thumbnail_detect
jetson-inference：Hello AI World指南，介绍如何使用TensorRT和NVIDIA Jetson部署...
1星 需积分: 49 86 浏览量
2021-02-03
05:47:59
上传
评论 1
收藏 71.25MB ZIP 举报 
jetson-inference：Hello AI World指南，介绍如何使用TensorRT和NVIDIA Jetson部署深度学习推理网络和深度视觉原语 （1612个子文件） 
resnet18_1025x321_net.cpp 59KB resnet18_2D_513x257_net.cpp 43KB tensorNet.cpp 43KB detectNet.cpp 40KB segNet.cpp 36KB PyDetectNet.cpp 26KB nvsmall_1025x321_net.cpp 21KB nvtiny_513x161_net.cpp 21KB PySegNet.cpp 19KB imageNet.cpp 18KB conv3d_transpose_plugin.cpp 16KB depthNet.cpp 16KB stereoNet.cpp 15KB conv3d_plugin.cpp 15KB homographyNet.cpp 14KB softargmax_plugin.cpp 12KB PyImageNet.cpp 12KB internal_utils.cpp 11KB flowNet.cpp 10KB FlattenConcat.cpp 9KB seg-img-tool.cpp 9KB odometry-plot.cpp 9KB depthWindow.cpp 9KB segnet.cpp 8KB stereonet-camera.cpp 8KB elu_plugin.cpp 7KB cost_volume_plugin.cpp 7KB PyTensorNet.cpp 7KB trt-bench.cpp 6KB odometryNet.cpp 6KB transform_plugin.cpp 6KB depthnet-camera.cpp 6KB flownet-camera.cpp 6KB homography-camera.cpp 5KB stereonet-console.cpp 5KB depthnet-console.cpp 5KB detectnet.cpp 5KB flownet-console.cpp 5KB imagenet.cpp 5KB superResNet.cpp 4KB depth-viewer.cpp 4KB homography-console.cpp 4KB conv_utils.cpp 4KB padding_plugin.cpp 4KB odometry-console.cpp 4KB superres-console.cpp 4KB slice_plugin.cpp 4KB my-recognition.cpp 3KB randInt8Calibrator.cpp 3KB PyInference.cpp 3KB intreleavingPlugin.cpp 2KB upsamplingPlugin.cpp 1KB trt-console.cpp 1KB slicePlugin.cpp 1KB doxygen.css 25KB doxygen.css 25KB tabs.css 8KB search.css 4KB navtree.css 2KB kernels.cu 13KB tensorConvert.cu 11KB segNet.cu 7KB superResNet.cu 5KB detectNet.cu 5KB odometryNet.cu 4KB homographyNet.cu 3KB upsampling.cu 2KB interleaving.cu 2KB stridedSlice.cu 2KB Dockerfile 3KB .editorconfig 76B
pytorch-capture-brontosaurus.gif 1.87MB .gitignore 60B .gitkeep 0B .gitmodules 686B detectNet.h 25KB tensorNet.h 23KB segNet.h 17KB imageNet.h 10KB depthNet.h 9KB upsamplingPlugin.h 7KB odometryNet.h 7KB flowNet.h 7KB interleavingPlugin.h 7KB slicePlugin.h 6KB internal_utils.h 6KB redtail_tensorrt_plugins.h 6KB stereoNet.h 6KB homographyNet.h 6KB conv_utils.h 5KB doc-groups.h 5KB tensorConvert.h 3KB PyInference.h 3KB superResNet.h 2KB randInt8Calibrator.h 2KB depthWindow.h 2KB networks.h 2KB PyTensorNet.h 2KB internal_macros.h 1KB PyDetectNet.h 1KB共 1612 条
- 1
- 2
- 3
- 4
- 5
- 6
- 17
评论1