19 KiB

原始文件 文件历史

Perception Tutorial

Phase 3: Cloud

In this phase of the tutorial, we will learn how to run our Scene on Unity Simulation and analyze the generated dataset using Dataset Insights. Unity Simulation will allow us to generate a much larger dataset than what is typically plausible on a workstation computer.

Steps included this phase of the tutorial:

Step 1: Setup Unity Account, Unity Simulation, and Cloud Project
Step 2: Run Project on Unity Simulation
Step 3: Keep Track of Your Runs Using the Unity Simulation Command-Line Interface
Step 4: Analyze the Dataset using Dataset Insights

Step 1: Setup Unity Account, Unity Simulation, and Cloud Project

In order to use Unity Simulation, you need to first create a Unity account or login with your existing one. Once logged in, you will also need to sign-up for Unity Simulation.

🟢 Action Click on the Cloud button at the top-right corner of Unity Editor to open the Services tab.

If you have not logged in yet, the Services tab will display a message noting that you are offline:

🟢 Action: Click Sign in... and follow the steps in the window that opens to sign in or create an account.
🟢 Action: Sign up for a free trial of Unity Simulation here.

Unity Simulation is a cloud-based service that makes it possible for you to run hundreds of instances of Unity builds in order to generate massive amounts of data. The Unity Simulation service is billed on a per-usage basis, and the free trial offers up to $100 of free credit per month. In order to access the free trial, you will need to provide credit card information. This information will be used to charge your account if you exceed the $100 monthly credit. A list of hourly and daily rates for various computational resources is available in the page where you first register for Unity Simulation.

Once you have registered for a free trial, you will be taken to your Unity Simulation dashboard, where you will be able to observe your usage and billing invoices.

It is now time to connect your local Unity project to a cloud project.

🟢 Action: Return to Unity Editor. In the Services tab click Select Organization and choose the only available option (which typically has the same name as your Unity username).

If you have used Unity before, you might have set-up multiple organizations for your account. In that case, choose whichever you would like to associate with this project.

🟢 Action: Click Create to create a new cloud project and connect your local project to it.

Step 2: Run Project on Unity Simulation

The process of running a project on Unity Simulation involves building it for Linux and then uploading this build, along with a set of parameters, to Unity Simulation. The Perception package simplifies this process by including a dedicated Run in Unity Simulation window that accepts a small number of required parameters and handles everything else automatically.

For performance reasons, it is best to disable real-time visualizations before carrying on with the Unity Simulation run.

🟢 Action: From the Inspector view of Perception Camera, disable real-time visualizations.

In order to make sure our builds are compatible with Unity Simulation, we need to set our project's scripting backend to Mono rather than IL2CPP (if not already set). We will also need to switch to Windowed mode.

🟢 Action: From the top menu bar, open Edit -> Project Settings.
🟢 Action: In the window that opens, navigate to the Player tab, find the Scripting Backend setting (under Other Settings), and change it to Mono:

🟢 Action: Change Fullscreen Mode to Windowed and set a width and height of 800 by 600.

🟢 Action: Close Project Settings.
🟢 Action: From the top menu bar, open Window -> Run in Unity Simulation.

🟢 Action: Choose TutorialScene (which is the Scene we have been working in) as your Main Scene and the SimulationScenario object as your Scenario.

Here, you can also specify a name for the run, the number of iterations the Scenario will execute for, and the number of Instances (number of nodes the work will be distributed across) for the run.

🟢 Action: Name your run FirstRun, set the number of iterations to 1000, and instances to 20.
🟢 Action: Click Build and Run.

Your project will now be built and then uploaded to Unity Simulation. Depending on the upload speed of your internet connection, this might take anywhere from a few seconds to a couple of minutes.

🟢 Action: Once the operation is complete, you can find the Build ID, Run Definition ID, and Execution ID of this Unity Simulation run in the Console tab:

Step 3: Keep Track of Your Runs Using the Unity Simulation Command-Line Interface

To keep track of the progress of your Unity Simulation run, you will need to use Unity Simulation's command-line interface (CLI). Detailed instructions for this CLI are provided here. For the purposes of this tutorial, we will only go through the most essential commands, which will help us know when our Unity Simulation run is complete and where to find the produced dataset.

🟢 Action: Download the latest version of unity_simulation_bundle.zip from here.

ℹ️ If you are using a MacOS computer, we recommend using the curl command from the Terminal to download the file, in order to avoid issues caused by the MacOS Gatekeeper when using the CLI. You can use these commands:

curl -Lo ~/Downloads/unity_simulation_bundle.zip <URL-unity_simulation_bundle.zip>
unzip ~/Downloads/unity_simulation_bundle.zip -d ~/Downloads/unity_simulation_bundle

The <URL-unity_simulation_bundle.zip> address can be found at the same page linked above.

🟢 Action: Extract the zip archive you downloaded.
🟢 Action: Open a command-line interface (Terminal on MacOS, cmd on Windows, etc.) and navigate to the extracted folder.

If you downloaded the zip archive in the default location in your downloads folder, you can use these commands to navigate to it from the command-line:

MacOS: cd ~/Downloads/unity_simulation_bundle

Windows: cd C:\Users\UserName\Downloads\unity_simulation_bundle

You will now be using the usim executable to interact with Unity Simulation through commands.

🟢 Action To see a list of available commands, simply run usim once:

MacOS: USimCLI/mac/usim

Windows: USimCLI\windows\usim

The first step is to login.

🟢 Action: Login to Unity Simulation using the usim login auth command.

MacOS: USimCLI/mac/usim login auth

Windows: USimCLI\windows\usim login auth

This command will ask you to press Enter to open a browser for you to login to your Unity account:

Press [ENTER] to open your browser to ...

🟢 Action: Press Enter to open a browser window for logging in.

Once you have logged you will see this page:

⚠️ On MacOS, you might get errors related to permissions. If that is the case, modify the permissions on the ~/.usim folder and its contents to give your user full read and write permission.

ℹ️ From this point on we will only include MacOS formatted commands in the tutorial, but all the usim commands we use will work in all supported operating systems.**

🟢 Action: Return to your command-line interface. Get a list of cloud projects associated with your Unity account using the usim get projects command:

MacOS: USimCLI/mac/usim get projects

Example output:

 name                  id                                       creation time             
--------------------- ---------------------------------------- --------------------------- 
 Perception Tutorial   acd31956-582b-4138-bec8-6670be150f09 *   2020-09-30T00:33:41+00:00 
 SynthDet              9ec23417-73cd-becd-9dd6-556183946153     2020-08-12T19:46:20+00:00

In case you have more than one cloud project, you will need to "activate" the one corresponding with your perception tutorial project. If there is only one project, it is already activated, and you will not need to execute the command below (note: replace <project-id> with the id of your desired project).

🟢 Action: Activate the relevant project:

MacOS: USimCLI/mac/usim activate project <project-id>

When asked if you are sure you want to change the active project, enter "y" and press Enter.

Now that we have made sure the correct project is active, we can get a list of all the current and past runs for the project.

🟢 Action: Use the usim get runs command to obtain a list of current and past runs:

MacOS: USimCLI/mac/usim get runs

An example output with 3 runs would look like this:

Active Project ID: acd31956-582b-4138-bec8-6670be150f09
name        id        creation time         executions                                    
----------- --------- --------------------- -----------------------------------------------
 FirstRun    1tLbZxL   2020-10-01 23:17:50    id        status        created_at           
                                             --------- ------------- --------------------- 
                                              yegz4WN   In_Progress   2020-10-01 23:17:54  
 Run2        klvfxgT   2020-10-01 21:46:39    id        status        created_at           
                                             --------- ------------- --------------------- 
                                              kML3i50   In_Progress   2020-10-01 21:46:42  
 Test        4g9xmW7   2020-10-01 02:27:06    id        status      created_at             
                                             --------- ----------- ---------------------   
                                              xBv3arj   Completed   2020-10-01 02:27:11

As seen above, each run has a name, an ID, a creation time, and a list of executions. Note that each "run" can have more than one "execution", as you can manually execute runs again using the CLI.

You can also obtain a list of all the builds you have uploaded to Unity Simulation using the usim get builds command.

You may notice that the IDs seen above for the run named FirstRun match those we saw earlier in Unity Editor's Console. You can see here that the single execution for our recently uploaded build is In_Progress and that the execution ID is yegz4WN.

Unity Simulation utilizes the ability to run simulation instances in parallel. If you enter a number larger than 1 for the number of instances in the Run in Unity Simulation window, your run will be parallelized, and multiple simulation instances will simultaneously execute. You can view the status of all simulation instances using the usim summarize run-execution <execution-id> command. This command will tell you how many instances have succeeded, failed, have not run yet, or are in progress. Make sure to replace <execution-id> with the execution ID seen in your run list. In the above example, this ID would be yegz4WN.

🟢 Action: Use the usim summarize run-execution <execution-id> command to observe the status of your execution nodes:

MacOS: USimCLI/mac/usim summarize run-execution <execution-id>

Here is an example output of this command, indicating that there is only one node, and that the node is still in progress:

 state         count 
------------- -------
 Successes     0     
 In Progress   1     
 Failures      0     
 Not Run       0

At this point, we will need to wait until the execution is complete. Check your run with the above command periodically until you see a 1 for Successes and 0 for In Progress. Given the relatively small size of our Scenario (1,000 Iterations), this should take less than 5 minutes.

🟢 Action: Use the usim summarize run-execution <execution-id> command periodically to check the progress of your run.
🟢 Action: When execution is complete, use the usim download manifest <execution-id> command to download the execution's manifest:

MacOS: USimCLI/mac/usim download manifest <execution-id>

The manifest is a .csv formatted file and will be downloaded to the same location from which you execute the above command, which is the unity_simulation_bundle folder. This file does not** include actual data, rather, it includes links to the generated data, including the JSON files, the logs, the images, and so on.

🟢 Action: Open the manifest file to check it. Make sure there are links to various types of output and check a few of the links to see if they work.

Step 4: Analyze the Dataset using Dataset Insights

In order to download the actual data from your run, we will now use Dataset Insights again. This time though, we will utilize some of the lines that were commented in our previous use with locally generated data.

🟢 Action: Open the Dataset Insights Jupyter notebook again, using the command below:

docker run -p 8888:8888 -v <download path>/data:/data -t unitytechnologies/datasetinsights:latest

In the above command, replace <download path> with the location on your computer in which you wish to download your data.

Once the Docker image is running, the rest of the workflow is quite similar to what we did in Phase 1, with certain differences caused by the need to download the data from Unity Simulation.

🟢 Action: Open a web browser and navigate to http://localhost:8888 to open the Jupyter notebook.
🟢 Action: Navigate to the datasetinsights/notebooks folder and open Perception_Statistics.ipynb.
🟢 Action: In the data_root = /data/<GUID> line, the <GUID> part will be the location inside your <download path> where the data will be downloaded. Therefore, you can just remove it so as to have data downloaded directly to the path you previously specified:

The next few lines of code pertain to setting up your notebook for downloading data from Unity Simulation.

🟢 Action: In the block of code titled "Unity Simulation [Optional]", uncomment the lines that assign values to variables, and insert the correct values, based on information from your Unity Simulation run.

We have previously learned how to obtain the run_execution_id and project_id. You can remove the value already present for annotation_definition_id and leave it blank. What's left is the access_token.

🟢 Action: Return to your command-line interface and run the usim inspect auth command.

MacOS: USimCLI/mac/usim inspect auth

If you receive errors regarding authentication, your token might have timed out. Repeat the login step (usim login auth) to login again and fix this issue.

A sample output from usim inspect auth will look like below:

Protect your credentials. They may be used to impersonate your requests.
access token: Bearer 0CfQbhJ6gjYIHjC6BaP5gkYn1x5xtAp7ZA9I003fTNT1sFp
expires in: 2:00:05.236227
expired: False
refresh token: FW4c3YRD4IXi6qQHv3Y9W-rwg59K7k0Te9myKe7Zo6M003f.k4Dqo0tuoBdf-ncm003fX2RAHQ
updated: 2020-10-02 14:50:11.412979

The access_token you need for your Dataset Insights notebook is the access token shown by the above command, minus the 'Bearer ' part. So, in this case, we should input 0CfQbhJ6gjYIHjC6BaP5gkYn1x5xtAp7ZA9I003fTNT1sFp in the notebook.

🟢 Action: Copy the access token excluding the 'Bearer ' part to the corresponding field in the Dataset Insights notebook.

Once you have entered all the information, the block of code should look like the screenshot below (the actual values you input will be different):

🟢 Action: Continue to the next code block and run it to download all the metadata files from the generated dataset. This includes JSON files and logs but does not include images (which will be downloaded later).

You will see a progress bar while the data downloads:

The next couple of code blocks (under "Load dataset metadata") analyze the downloaded metadata and display a table containing annotation-definition-ids for the various metrics defined in the dataset.

🟢 Action: Once you reach the code block titled "Built-in Statistics", make sure the value assigned to the field rendered_object_info_definition_id matches the id displayed for this metric in the table output by the code block immediately before it. The screenshot below demonstrates this (note that your ids might differ from the ones here):

Follow the rest of the steps inside the notebook to generate a variety of plots and stats. Keep in mind that this notebook is provided just as an example, and you can modify and extend it according to your own needs using the tools provided by the Dataset Insights framework.

This concludes the Perception tutorial. The next step in this workflow would be to train an object-detection model using a dataset generated on Unity Simulation. It is important to note that the 1000 large dataset we generated here is probably not sufficiently large for training most models. We chose this number here so that the run would complete in a fairly short period of time, allowing us to move on to learning how to analyze the dataset's statistics. In order to generate data for training, we recommend a dataset of about 400,000 captures.

In the near future, we will expand this tutorial to Phase 4, which will include instructions on how to train a Faster R-CNN object-detection model using a dataset that can be generated by following this tutorial.

19 KiB 原始文件 文件历史