21 lines
6.2 KiB
Markdown
21 lines
6.2 KiB
Markdown
|
| Argument | Type | Default | Range | Description |
|
||
|
|
| ----------------------------------------------------------------------------------------- | ------- | ----------------------- | ------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
|
||
|
|
| [`hsv_h`](../guides/yolo-data-augmentation.md/#hue-adjustment-hsv_h) | `float` | `{{ hsv_h }}` | `0.0 - 1.0` | Adjusts the hue of the image by a fraction of the color wheel, introducing color variability. Helps the model generalize across different lighting conditions. |
|
||
|
|
| [`hsv_s`](../guides/yolo-data-augmentation.md/#saturation-adjustment-hsv_s) | `float` | `{{ hsv_s }}` | `0.0 - 1.0` | Alters the saturation of the image by a fraction, affecting the intensity of colors. Useful for simulating different environmental conditions. |
|
||
|
|
| [`hsv_v`](../guides/yolo-data-augmentation.md/#brightness-adjustment-hsv_v) | `float` | `{{ hsv_v }}` | `0.0 - 1.0` | Modifies the value (brightness) of the image by a fraction, helping the model to perform well under various lighting conditions. |
|
||
|
|
| [`degrees`](../guides/yolo-data-augmentation.md//#rotation-degrees) | `float` | `{{ degrees }}` | `0.0 - 180` | Rotates the image randomly within the specified degree range, improving the model's ability to recognize objects at various orientations. |
|
||
|
|
| [`translate`](../guides/yolo-data-augmentation.md/#translation-translate) | `float` | `{{ translate }}` | `0.0 - 1.0` | Translates the image horizontally and vertically by a fraction of the image size, aiding in learning to detect partially visible objects. |
|
||
|
|
| [`scale`](../guides/yolo-data-augmentation.md/#scale-scale) | `float` | `{{ scale }}` | `>=0.0` | Scales the image by a gain factor, simulating objects at different distances from the camera. |
|
||
|
|
| [`shear`](../guides/yolo-data-augmentation.md/#shear-shear) | `float` | `{{ shear }}` | `-180 - +180` | Shears the image by a specified degree, mimicking the effect of objects being viewed from different angles. |
|
||
|
|
| [`perspective`](../guides/yolo-data-augmentation.md/#perspective-perspective) | `float` | `{{ perspective }}` | `0.0 - 0.001` | Applies a random perspective transformation to the image, enhancing the model's ability to understand objects in 3D space. |
|
||
|
|
| [`flipud`](../guides/yolo-data-augmentation.md/#flip-up-down-flipud) | `float` | `{{ flipud }}` | `0.0 - 1.0` | Flips the image upside down with the specified probability, increasing the data variability without affecting the object's characteristics. |
|
||
|
|
| [`fliplr`](../guides/yolo-data-augmentation.md/#flip-left-right-fliplr) | `float` | `{{ fliplr }}` | `0.0 - 1.0` | Flips the image left to right with the specified probability, useful for learning symmetrical objects and increasing dataset diversity. |
|
||
|
|
| [`bgr`](../guides/yolo-data-augmentation.md/#bgr-channel-swap-bgr) | `float` | `{{ bgr }}` | `0.0 - 1.0` | Flips the image channels from RGB to BGR with the specified probability, useful for increasing robustness to incorrect channel ordering. |
|
||
|
|
| [`mosaic`](../guides/yolo-data-augmentation.md/#mosaic-mosaic) | `float` | `{{ mosaic }}` | `0.0 - 1.0` | Combines four training images into one, simulating different scene compositions and object interactions. Highly effective for complex scene understanding. |
|
||
|
|
| [`mixup`](../guides/yolo-data-augmentation.md/#mixup-mixup) | `float` | `{{ mixup }}` | `0.0 - 1.0` | Blends two images and their labels, creating a composite image. Enhances the model's ability to generalize by introducing label noise and visual variability. |
|
||
|
|
| [`cutmix`](../guides/yolo-data-augmentation.md/#cutmix-cutmix) | `float` | `{{ cutmix }}` | `0.0 - 1.0` | Combines portions of two images, creating a partial blend while maintaining distinct regions. Enhances model robustness by creating occlusion scenarios. |
|
||
|
|
| [`copy_paste`](../guides/yolo-data-augmentation.md/#copy-paste-copy_paste) | `float` | `{{ copy_paste }}` | `0.0 - 1.0` | _Segmentation only_. Copies and pastes objects across images to increase object instances. |
|
||
|
|
| [`copy_paste_mode`](../guides/yolo-data-augmentation.md/#copy-paste-mode-copy_paste_mode) | `str` | `{{ copy_paste_mode }}` | - | _Segmentation only_. Specifies the `copy-paste` strategy to use. Options include `'flip'` and `'mixup'`. |
|
||
|
|
| [`auto_augment`](../guides/yolo-data-augmentation.md/#auto-augment-auto_augment) | `str` | `{{ auto_augment }}` | - | _Classification only_. Applies a predefined augmentation policy (`'randaugment'`, `'autoaugment'`, or `'augmix'`) to enhance model performance through visual diversity. |
|
||
|
|
| [`erasing`](../guides/yolo-data-augmentation.md/#random-erasing-erasing) | `float` | `{{ erasing }}` | `0.0 - 0.9` | _Classification only_. Randomly erases regions of the image during training to encourage the model to focus on less obvious features. |
|