Through-out the tutorial, lines starting with bullet points followed by **"Action:"** denote the individual actions you will need to perform in order to progress through the tutorial. This is while non-bulleted lines will provide additional context and explanation around the actions. If in a hurry, you can just follow the actions!
Steps included this phase of the tutorial:
- [Download Unity Editor and Create a New Project](#step-1)
- [Download the Perception Package and Import Samples](#step-2)
- [Step 1: Download Unity Editor and Create a New Project](#step-1)
- [Step 2: Download the Perception Package and Import Samples](#step-2)
- [Step 3: Setup a Scene for Your Perception Simulation](#step-3)
- [Step 4: Specify Ground-Truth and Setup Object Labels](#step-4)
- [Step 5: Add and Set-up Randomizers](#step-5)
- [Step 6: Generate and Verify Synthetic Data](#step-6)
### Step 1: Download Unity Editor and Create a New Project {#step-1}
### <aname="step-1">Step 1: Download Unity Editor and Create a New Project</a>
* **Action**: Navigate to [this](https://unity3d.com/get-unity/download/archive) page to download and install the latest version of Unity Editor 2020.1.x. (The tutorial has not yet been tested on newer versions.)
An alternative approach is to first install [_**Unity Hub**_](https://unity3d.com/get-unity/download), which will allow you to have multiple versions of Unity on your computer, and make it easier to manage your Unity projects and the versions of Unity they will use.
When you first run Unity, you will be asked to open an existing project, or create a new one.
* **Action**: Open Unity and create a new project using the Universal Render Pipeline. Name your new project _**Perception Tutorial**_, as shown below.
* **Action**: Open Unity and create a new project using the Universal Render Pipeline. Name your new project _**Perception Tutorial**_, and specify a desired location as shown below.
Once your new project is created and loaded, you will be presented with the Unity Editor interface. From this point, whenever we refer to _the editor_, we mean Unity Editor.
Once your new project is created and loaded, you will be presented with the Unity Editor interface. From this point on, whenever we refer to _**the editor**_, we mean Unity Editor.
* **Action**: From the top menu bar, open _**Window**_ -> _**Package Manager**_.
As the name suggests, the _**Package Manager**_ is where you can download new packages, update or remove existing ones, and access a variety of information and additional actions for each package.
**Note:** If you would like a specific version of the package, you can append the version to the end of the url. For example `com.unity.perception@0.1.0-preview.5`. For this tutorial, **we do not need to add a version**. You can also install packages from a local clone of the Perception repository. For inofrmation on installing local pacakges is available [here](ttps://docs.unity3d.com/Manual/upm-ui-local.html).
**Note:** If you would like a specific version of the package, you can append the version to the end of the url. For example `com.unity.perception@0.1.0-preview.5`. For this tutorial, **we do not need to add a version**. You can also install the package from a local clone of the Perception repository. More inofrmation on installing local pacakges is available [here](https://docs.unity3d.com/Manual/upm-ui-local.html).
It will take some time for the manager to download and import the package. Once the operation finishes, you will see the newly download Perception package automatically selected in the _**Package Manager**_, as depicted below:
Each package can come with a set of samples. As seen in the righthand panel, the Perception package includes a sample named _**Tutorial Files**_, which will be required for completing this tutorial. The sample files consist of example foreground and background objects (foreground: objects that the eventual machine learning model will try to detect, background: objects that will be placed in the background as distractions to for the machine learning model), randomizers, shaders, and other useful elements to work with during this tutorial.
Each package can come with a set of samples. As seen in the righthand panel, the Perception package includes a sample named _**Tutorial Files**_, which will be required for completing this tutorial. The sample files consist of example foreground and background objects, randomizers, shaders, and other useful elements to work with during this tutorial. **Foreground** objects are those thatthe eventual machine learning model will try to detect, and **background** objects will be placed in the background as distractors for the model.
* **Action**: In the _**Package Manager**_ window, from the list of _**Samples**_ for the Perception package, click on the _**Import into Project**_ button for the sample named _**Tutorial Files**_.
This step prepares your project to render tailor-made images that will be later used for labeling the generates synthetic data.
### Step 3: Setup a Scene for Your Perception Simulation
### <aname="step-3">Step 3: Setup a Scene for Your Perception Simulation</a>
Simply put, in Unity, Scenes contain any object that exists in the world. This world can be a game, or in this case, a perception-oriented simulation. Every new project contains a Scene named `SampleScene`, which is automatically openned when the project is created. This Scenes comes with several objects and settings that we do not need, so let's create a new one.
* **Action**: In the _**Project**_ tab, right-click on the `Assets/Scenes` folder and click _**Create -> Scene**_. Name this new Scene `TutorialScene` and double-click on it to open it.
One of the useful features that comes with the `Perception Camera` is the ability to display real-time visualizations of the labelers when your simulation is running. For instance, the `BoundingBox2DLabeler` can display bounding boxes around the foreground objects that it tracks in real-time. To enable this feature, make sure the `Show Labeler Visualizations` checkmark is enabled.
### Step 4: Specify Ground-Truth and Object Labels
### <aname="step-4">Step 4: Specify Ground-Truth and Setup Object Labels</a>
It is now time to tell your each labeler added to the `Perception Camera` which objects it should label in the generated dataset. For instance, if your workflow is intended to generate frames and ground-truth for detecting chairs, your labelers would need to know that they should look for objects labeled "chair" within the scene. The chairs should in turn also be labeled "chair" in order to make them visible to their intended labelers. We will now learn how to set-up these configuartions.
Note that each object can have multiple labels assigned, and thus appear as different objects to labelers with different label configurations. For instance, you may want your semantic segmentation labeler to detect all cream cartons as as `dairy_product`, while your bounding box labeler still distinguishes between different types of dairy product. To achieve this, you can add a `dairy_product` label to all your dairy products, and then in your label configuration for semantic segmentation, only add the `dairy_product` label, and not any specific products or brand names. To add an additional a label to the cream carton, you can click on the _**+**_ button to the bottom right corner of the label list, in the `Labeling` component.
### Step 5: Add and Set-up Randomizers
### <aname="step-5">Step 5: Add and Set-up Randomizers</a>
As mentioned earlier, one of the core ingredients of the perception workflow is the randomization of various aspects of the simulation, in order to introduce sufficient variation into the generated data.
* **Action**: Drag `ForegroundObjectPlacementRandomizer` and drop it above `RotationRandomizer`.
### Step 6: Generate and Verify Synthetic Data
### <aname="step-6">Step 6: Generate and Verify Synthetic Data</a>
You are now ready to generate your first dataset. The `FixedLengthScenario` component that we used has 1000 `Iteration`s by default, and each `Iteration` runs for one frame. Thus, the simulation will produce 1000 frames of annotated captures.
GitHub
4 年前