Implement Omegaconfig PR1: basic functionality

sleap_nn/config/data_config.py

@@ -0,0 +1,143 @@
+import attrs
+from omegaconf import MISSING
+from typing import Optional, Tuple, List, Dict
+
+
+"""Serializable configuration classes for specifying all data configuration parameters.
+
+These configuration classes are intended to specify all 
+the parameters required to initialize the data config.
+"""
+
+
+@attrs.define
+class DataConfig:
+    """Data configuration.
+
+    provider: (str) Provider class to read the input sleap files. Only "LabelsReader" supported for the training pipeline.
+    train_labels_path: (str) Path to training data (.slp file)
+    val_labels_path: (str) Path to validation data (.slp file)
+    user_instances_only: (bool) True if only user labeled instances should be used for training. If False, both user labeled and predicted instances would be used. Default: True.
+    chunk_size: (int) Size of each chunk (in MB). Default: "100". #TODO: change in inference ckpts
+    preprocessing: Configuration options related to data preprocessing.
+    use_augmentations_train: (bool) True if the data augmentation should be applied to the training data, else False.
+    augmentation_config: Configurations related to augmentation
+    instance_cropping: Configuration options related to instance cropping for centroid
+        and topdown models.
+    """
+
+    provider: str = "LabelsReader"
+    train_labels_path: str = MISSING
+    val_labels_path: str = MISSING
+    user_instances_only: bool = True
+    chunk_size: int = 100
+    preprocessing: PreprocessingConfig = attrs.field(factory=PreprocessingConfig)
+    use_augmentations_train: bool = False
+    augmentation_config: Optional[AugmentationConfig] = None
+
+
+@attrs.define
+class PreprocessingConfig:
+    """Configuration of Preprocessing.
+
+    Attributes:
+        is_rgb: (bool) True if the image has 3 channels (RGB image). If input has only one channel when this is set to True, then the images from single-channel is replicated along the channel axis. If input has three channels and this is set to False, then we convert the image to grayscale (single-channel) image.
+        max_height: (int) Maximum height the image should be padded to. If not provided, the original image size will be retained. Default: None.
+        max_width: (int) Maximum width the image should be padded to. If not provided, the original image size will be retained. Default: None.
+        scale: (float or List[float]) Factor to resize the image dimensions by, specified as either a float scalar or as a 2-tuple of [scale_x, scale_y]. If a scalar is provided, both dimensions are resized by the same factor.
+        crop_hw: (Tuple[int]) Crop height and width of each instance (h, w) for centered-instance model. If None, this would be automatically computed based on the largest instance in the sio.Labels file.
+        min_crop_size: (int) Minimum crop size to be used if crop_hw is None.
+    """
+
+    is_rgb: bool = True
+    max_height: Optional[int] = None
+    max_width: Optional[int] = None
+    scale: Union[float, Tuple[float, float]] = 1.0
+    crop_hw: Optional[Tuple[int, int]] = None
+    min_crop_size: int = 32  # to help app work incase of error
+
+
+@attrs.define
+class AugmentationConfig:
+    """Configuration of Augmentation
+
+    Attributes:
+        random crop: (Optional) (Dict[float]) {"random_crop_p": None, "crop_height": None. "crop_width": None}, where random_crop_p is the probability of applying random crop and crop_height and crop_width are the desired output size (out_h, out_w) of the crop.
+        intensity: (Optional)
+        geometric: (Optional)
+    """
+
+    random_crop: Optional[Dict[str, Optional[float]]] = None
+    intensity: Optional[IntensityConfig] = attrs.field(default=None)
+    geometric: Optional[GeometricConfig] = attrs.field(default=None)
+
+
+@attrs.define
+class IntensityConfig:
+    """Configuration of Intensity (Optional):
+
+    Attributes:
+        uniform_noise_min: (float) Minimum value for uniform noise (uniform_noise_min >=0).
+        uniform_noise_max: (float) Maximum value for uniform noise (uniform_noise_max <>=1).
+        uniform_noise_p: (float) Probability of applying random uniform noise. Default=0.0
+        gaussian_noise_mean: (float) The mean of the gaussian noise distribution.
+        gaussian_noise_std: (float) The standard deviation of the gaussian noise distribution.
+        gaussian_noise_p: (float) Probability of applying random gaussian noise. Default=0.0
+        contrast_min: (float) Minimum contrast factor to apply. Default: 0.5.
+        contrast_max: (float) Maximum contrast factor to apply. Default: 2.0.
+        contrast_p: (float) Probability of applying random contrast. Default=0.0
+        brightness: (float) The brightness factor to apply. Default: (1.0, 1.0).
+        brightness_p: (float) Probability of applying random brightness. Default=0.0
+    """
+
+    uniform_noise_min: float = attrs.field(
+        default=0.0, validator=attrs.validators.ge(0)
+    )
+    uniform_noise_max: float = attrs.field(
+        default=1.0, validator=attrs.validators.le(1)
+    )
+    uniform_noise_p: float = 0.0
+    gaussian_noise_mean: float = 0.0
+    gaussian_noise_std: float = 1.0
+    gaussian_noise_p: float = 0.0
+    contrast_min: float = attrs.field(default=0.5, validator=attrs.validators.ge(0))
+    contrast_max: float = attrs.field(default=2.0, validator=attrs.validators.ge(0))
+    contrast_p: float = 0.0
+    brightness: Tuple[float, float] = (1.0, 1.0)
+    brightness_p: float = 0.0
+
+
+@attrs.define
+class GeometricConfig:
+    """
+    Configuration of Geometric (Optional)
+
+    Attributes:
+        rotation: (float) Angles in degrees as a scalar float of the amount of rotation. A random angle in (-rotation, rotation) will be sampled and applied to both images and keypoints. Set to 0 to disable rotation augmentation.
+        scale: (float) scaling factor interval. If (a, b) represents isotropic scaling, the scale is randomly sampled from the range a <= scale <= b. If (a, b, c, d), the scale is randomly sampled from the range a <= scale_x <= b, c <= scale_y <= d Default: None.
+        translate_width: (float) Maximum absolute fraction for horizontal translation. For example, if translate_width=a, then horizontal shift is randomly sampled in the range -img_width * a < dx < img_width * a. Will not translate by default.
+        translate_height: (float) Maximum absolute fraction for vertical translation. For example, if translate_height=a, then vertical shift is randomly sampled in the range -img_height * a < dy < img_height * a. Will not translate by default.
+        affine_p: (float) Probability of applying random affine transformations. Default=0.0
+        erase_scale_min: (float) Minimum value of range of proportion of erased area against input image. Default: 0.0001.
+        erase_scale_max: (float) Maximum value of range of proportion of erased area against input image. Default: 0.01.
+        erase_ration_min: (float) Minimum value of range of aspect ratio of erased area. Default: 1.
+        erase_ratio_max: (float) Maximum value of range of aspect ratio of erased area. Default: 1.
+        erase_p: (float) Probability of applying random erase. Default=0.0
+        mixup_lambda: (float) min-max value of mixup strength. Default is 0-1. Default: None.
+        mixup_p: (float) Probability of applying random mixup v2. Default=0.0
+        input_key: (str) Can be image or instance. The input_key instance expects the KorniaAugmenter to follow the InstanceCropper else image otherwise for default.
+    """
+
+    rotation: float = 0.0
+    scale: Optional[Tuple[float, float, float, float]] = None
+    translate_width: float = 0.0
+    translate_height: float = 0.0
+    affine_p: float = 0.0
+    erase_scale_min: float = 0.0001
+    erase_scale_max: float = 0.01
+    erase_ratio_min: float = 1.0
+    erase_ratio_max: float = 1.0
+    erase_p: float = 0.0
+    mixup_lambda: Optional[float] = None
+    mixup_p: float = 0.0
+    input_key: str = "image"