PoseAtlas 图库:
PoseAtlasLoader is a versatile node designed to facilitate the selection and retrieval of images from the RuiC-PoseAtlas library, which encompasses a diverse collection of poses, costumes, and portraits. This node allows you to access a comprehensive database of 4,345 entries, each meticulously categorized and described. The PoseAtlasLoader node is particularly beneficial for AI artists seeking to incorporate detailed and varied visual references into their projects. By providing options to select different collections and rendering styles, it enables you to tailor the output to your specific artistic needs, whether you're looking for line sketches, grayscale depth maps, or full-color images. The node's functionality is seamlessly integrated into the ComfyUI environment, ensuring a smooth and efficient workflow for accessing and utilizing these rich visual resources.
PoseAtlas 图库 Input Parameters:
collection
The collection parameter allows you to choose from three distinct categories within the RuiC-PoseAtlas library: poses, costumes, and portraits. Each category offers a unique set of images tailored to different artistic needs. The default value is "姿势" (poses), which includes 1,439 entries. Selecting a different collection will require you to reselect the specific item you wish to use. This parameter is crucial for narrowing down the type of visual reference you need for your project.
pose
The pose parameter lets you select a specific image from the chosen collection. This parameter is populated with labels from the entire library, allowing you to choose the exact visual reference you need. The default value is the first label in the list, if available. This selection directly impacts the image that will be loaded and used in your project, making it essential for obtaining the desired visual output.
render
The render parameter determines the rendering style of the image, specifically for the poses collection. You can choose between "线稿" (line sketch) and "深度" (depth map). The default is "线稿," which provides a line sketch of the pose. The depth map option offers a grayscale image where white areas are closer to the camera. This parameter is not applicable to the costumes and portraits collections, which only have original images.
invert_mask
The invert_mask parameter is a boolean option that allows you to invert the mask of the image. By default, this is set to False, meaning the mask is not inverted. When set to True, the mask will be inverted, which can be useful for certain compositional techniques or effects where the inverse of the mask is required.
PoseAtlas 图库 Output Parameters:
image
The image output provides the selected visual reference as an image tensor. This output is crucial for integrating the chosen image into your AI art project, allowing you to manipulate or analyze it further within the ComfyUI environment.
mask
The mask output delivers a mask tensor corresponding to the selected image. This mask can be used for various purposes, such as isolating specific parts of the image or applying effects selectively. The mask is particularly useful when working with images that have transparency or require compositional adjustments.
item_name
The item_name output is a string that describes the selected image, including its category and specific attributes. This output is valuable for keeping track of the images used in your project, providing a clear reference to the source and characteristics of each visual element.
PoseAtlas 图库 Usage Tips:
- Ensure that you have downloaded the necessary images by running the provided download script before using the node to avoid missing file errors.
- Utilize the
renderparameter to switch between line sketches and depth maps for poses, depending on the visual style you aim to achieve in your project.
PoseAtlas 图库 Common Errors and Solutions:
PoseAtlas: 找不到条目 '<pose>'
- Explanation: This error occurs when the specified pose cannot be found in the library.
- Solution: Double-check the pose name for typos or ensure that the correct collection is selected.
PoseAtlas: 图片未下载 —— <path>
- Explanation: This error indicates that the image file has not been downloaded to your local system.
- Solution: Run the
python scripts/download_images.pyscript in the node directory to download the necessary images, then restart ComfyUI.
