Using Commands from the Jetson-Inference Library

If you want to repeat the commands from the instructor's demonstration, you can use these steps and commands to get the same results.

ImageNet Demo

There are multiple types of deep learning networks available, including recognition, detection/localization, and semantic segmentation. The first deep learning capability we're highlighting in this tutorial is image recognition, using classifcation networks that have been trained on large datasets to identify scenes and objects.

The imageNet object accepts an input image and outputs the probability for each class. Having been trained on the ImageNet ILSVRC dataset of 1000 objects, the GoogleNet and ResNet-18 models were automatically downloaded during the build step. See below for other classification models that can be downloaded and used as well.

From the terminal inside VSCode, run this command:
```
imagenet data/images/fruit_7.jpg /fruit_test.jpg
```

Download this video:

wget https://nvidia.box.com/shared/static/tlswont1jnyu3ix2tbf7utaekpzcx4rc.mkv -O jellyfish.mkv

Run the video through imagenet to see it classified.

imagenet --network=resnet-18 jellyfish.mkv jellyfish_resnet18.mkv

Download the video to view it on your host machine.

You are welcome to do some more if you like, or move into the next network.

DetectNet Demo

The previous recognition examples output class probabilities representing the entire input image. Next, you're going to focus on object detection, and finding where in the frame various objects are located by extracting their bounding boxes. Unlike image classification, object detection networks are capable of detecting many different objects per frame. Object detection models are machine learning networks that go through images and detect objects in them. Follow these steps to run the detection network on your Nano.

Demonstration of detectnet.

You can also test detectNet on any images or videos using the same arguments that you used with imageNet.

Learn more about Object Detection here.

From the terminal inside VSCode, run this command:
```
detectnet data/images/human2.jpg /humans_test.jpg
```

Download this video:

wget https://nvidia.box.com/shared/static/veuuimq6pwvd62p9fresqhrrmfqz0e2f.mp4 -O pedestrians.mp4

Run the video through detectnet to see what the network detects.
```
detectnet pedestrians.mp4 /pedestrians_ssd.mp4
```
Download the video to view it on your host machine.

You are welcome to do some more if you like, or move into the next network.

SegNet Demo

In addition to object detection, the nano also had the ability to do semantic segmentation. Semantic segmentation is based on image recognition, except the classifications occur at the pixel level as opposed to the entire image. This is accomplished by convolutionalizing a pre-trained image recognition backbone, which transforms the model into a Fully Convolutional Network (FCN) capable of per-pixel labeling. Especially useful for environmental perception, segmentation yields dense per-pixel classifications of many different potential objects per scene, including scene foregrounds and backgrounds.

Demonstration of segnet.

While it's running, you can stop it at any time using Ctl+C. You can also test segNet on any images or videos using the same arguments that you used with imageNet and detectNet.

Learn more about semantic segmentation here.

Here's an example of segmenting an urban street scene with the Cityscapes model:
```
segnet --network=fcn-resnet18-cityscapes images/city_0.jpg output.jpg
```

You are welcome to do some more if you like, or move into the next network.

PoseNet Demo

Pose estimation consists of locating various body parts (aka keypoints) that form a skeletal topology (aka links). Pose estimation has a variety of applications including gestures, AR/VR, HMI (human/machine interface), and posture/gait correction. Pre-trained models are provided for human body and hand pose estimation that are capable of detecting multiple people per frame.

The poseNet object accepts an image as input, and outputs a list of object poses. Each object pose contains a list of detected keypoints, along with their locations and links between keypoints.

From the terminal inside VSCode, run this command:

posenet "images/humans_*.jpg" pose_humans_%i.jpg

Use your webcam to record a short video of your hand moving in front of the camera.
```
posenet --network=resnet18-hand /dev/video0 output.mp4
```
Download the video to view it on your host machine.

You are welcome to do some more if you like, or move into the next network.