A README for the preprocessing layers (keras-team#1979)

* init

* nits

* review comments

* nits

keras_cv/layers/preprocessing/README.md

@@ -0,0 +1,54 @@
+# Preprocessing Layers
+
+KerasCV offers many preprocessing and data augmentation layers which support classification, object detection, and segmentation masks. When you use KerasCV augmentation layers to augment your training data, class labels, bounding boxes, and mask labels automatically get augmented alongside the image augmentations!
+
+The provided table gives an overview of the different augmentation layers available and the data formats they support.
+
+| Layer Name | Vectorized | Segmentation Masks | BBoxes | Class Labels |
+| :-- | :--: | :--: | :--: | :--: |
+| AugMix | ❌ | ❌ | ✅ | ✅ |
+| AutoContrast | ✅ | ✅ | ✅ | ✅ |
+| ChannelShuffle | ✅ | ✅ | ✅ | ✅ |
+| CutMix | ❌ | ✅ | ❌ | ✅ |
+| Equalization | ❌ | ✅ | ✅ | ✅ |
+| FourierMix | ❌ | ❌ | ❌ | ✅ |
+| Grayscale | ✅ | ✅ | ✅ | ✅ |
+| GridMask | ❌ | ✅ | ✅ | ✅ |
+| JitteredResize | ✅ | ✅ | ✅ | ✅ |
+| MixUp | ❌ | ✅ | ✅ | ✅ |
+| Mosaic | ✅ | ✅ | ✅ | ✅ |
+| Posterization | ❌ | ✅ | ✅ | ✅ |
+| RandAugment | ❌ | ❌ | ❌ | ❌ |
+| RandomApply <sup>+</sup> | - | - | - | - |
+| RandomAspectRatio | ❌ | ❌ | ✅ | ✅ |
+| RandomBrightness | ✅| ✅ | ✅ | ✅ |
+| RandomChannelShift | ❌| ✅ | ✅ | ✅ |
+| RandomChoice <sup>+</sup> | - | - | - | - |
+| RandomColorDegeneration | ❌ | ✅ | ✅ | ✅ |
+| RandomColorJitter | ✅ | ✅ | ✅ | ✅ |
+| RandomContrast | ✅ | ✅ | ✅ | ✅ |
+| RandomCropAndResize | ❌ | ✅ | ✅ | ❌ |
+| RandomCrop | ✅ | ❌ | ✅ | ✅ |
+| RandomCutout | ❌ | ❌ | ❌ | ✅ |
+| RandomFlip | ✅ | ✅ | ✅ | ✅ |
+| RandomGaussianBlur | ❌ | ✅ | ✅ | ✅ |
+| RandomHue | ✅ | ✅ | ✅ | ✅ |
+| RandomJpegQuality | ❌ | ✅ | ✅ | ✅ |
+| RandomRotation | ✅ | ✅ | ✅ | ✅ |
+| RandomSaturation | ✅ | ✅ | ✅ | ✅ |
+| RandomSharpness | ✅ | ✅ | ✅ | ✅ |
+| RandomShear | ✅ | ❌ | ✅ | ✅ |
+| RandomTranslation | ✅ | ❌ | ✅ | ✅ |
+| RandomZoom | ✅ | ❌ | ❌ | ✅ |
+| RepeatedAugmentation <sup>+</sup> | - | - | - | - |
+| Rescaling | ❌ | ✅ | ✅ | ✅ |
+| Resizing | ❌ | ✅ | ✅ | ❌ |
+| Solarization | ✅ | ✅ | ✅ | ✅ |
+
+<sup>+</sup> Meta Layers, the data types will depend on the Sub Layers.
+
+# Base Layers
+
+- BaseImageAugmentationLayer
+- VectorizedBaseImageAugmentationLayer
+- RandomAugmentationPipeline

keras_cv/layers/preprocessing/base_image_augmentation_layer.py

@@ -76,7 +76,7 @@ class BaseImageAugmentationLayer(base_class):
     The `call()` method supports two formats of inputs:
     1. A single image tensor with shape (height, width, channels) or
        (batch_size, height, width, channels)
-    1. A dict of tensors with any of the following keys (note that `"images"`
+    2. A dict of tensors with any of the following keys (note that `"images"`
        must be present):
        * `"images"` - Image Tensor with shape (height, width, channels) or
         (batch_size, height, width, channels)