cg-image-picker
cg-image-picker is a node designed to allow users to preview images within a workflow and select one or more to pass on to subsequent stages, either as images or latents. It offers various modes for handling image selection, including always pausing for user selection, repeating the last selection, and passing through all images, among others, to accommodate different use cases.
Repo info
- Repo url:
https://github.com/chrisgoringe/cg-image-picker - Commit hash:
e1c88898865072139d8a8eaa9dab1ed19dc73a2f
Licenses
For this pack, no licenses were identified
Preview Chooser Fabric
Documentation
- Class name:
Preview Chooser Fabric - Category:
image_chooser - Output node:
False
This node extends the functionality of image selection by allowing users to preview and choose between images based on their latent representations. It supports advanced selection mechanisms, including the ability to differentiate between positive and negative selections within a batch of images, thereby facilitating more nuanced decision-making processes in workflows that involve image analysis and manipulation.
Input types
Required
images- The images parameter is crucial for providing the visual content that users will preview and select from. It directly influences the node's ability to render previews and capture user choices.
- Comfy dtype:
IMAGE - Python dtype:
torch.Tensor
latents- The latents parameter represents the latent representations of the images provided. It is essential for the node to perform operations on the images based on their underlying characteristics, enabling more sophisticated selection criteria.
- Comfy dtype:
LATENT - Python dtype:
torch.Tensor
Output types
positive- Comfy dtype:
LATENT - The positive output represents the selected images or latents deemed favorable by the user, enabling downstream processes to utilize these preferred choices.
- Python dtype:
torch.Tensor
- Comfy dtype:
negative- Comfy dtype:
LATENT - The negative output contrasts with the positive by representing the selections marked as unfavorable, allowing for the exclusion or alternative handling of these choices in subsequent operations.
- Python dtype:
torch.Tensor
- Comfy dtype:
Usage tips
- Infra type:
GPU - Common nodes: unknown
Preview Chooser
Documentation
- Class name:
Preview Chooser - Category:
image_chooser - Output node:
False
The Preview Chooser node is designed to display a set of images to the user, allowing them to visually inspect and select one or more images to proceed with in the workflow. This node integrates a user interaction step into the data processing pipeline, facilitating decision-making based on visual information.
Input types
Required
mode- The 'mode' parameter determines the operational mode of the node, affecting how user selections are processed and how the node behaves in different scenarios.
- Comfy dtype:
COMBO[STRING] - Python dtype:
str
count- The 'count' parameter specifies the number of images to select or process, allowing for control over the quantity of output based on user input or predefined settings.
- Comfy dtype:
INT - Python dtype:
int
Optional
images- The 'images' parameter represents the collection of images to be displayed for user selection. It plays a crucial role in the node's operation by providing the visual content that users interact with to make their selections.
- Comfy dtype:
IMAGE - Python dtype:
torch.Tensor
latents- The 'latents' parameter refers to the latent representations associated with the images. It is used to carry forward the selected images' latent information through the workflow, enabling further processing or analysis.
- Comfy dtype:
LATENT - Python dtype:
torch.Tensor
masks- The 'masks' parameter provides optional mask information for the images, which can be used for advanced selection or processing scenarios where specific parts of images are of interest.
- Comfy dtype:
MASK - Python dtype:
Optional[torch.Tensor]
Output types
images- Comfy dtype:
IMAGE - Returns the images selected by the user.
- Python dtype:
torch.Tensor
- Comfy dtype:
latents- Comfy dtype:
LATENT - Returns the latent representations of the selected images.
- Python dtype:
torch.Tensor
- Comfy dtype:
masks- Comfy dtype:
MASK - Returns the mask information for the selected images, if applicable.
- Python dtype:
Optional[torch.Tensor]
- Comfy dtype:
selected- Comfy dtype:
STRING - Provides a string or identifier for the selected images, facilitating tracking and further processing.
- Python dtype:
str
- Comfy dtype:
Usage tips
- Infra type:
CPU - Common nodes:
- ImageBlend
- Mute / Bypass Repeater (rgthree)
- PreviewImage
- IterativeImageUpscale
- ImageUpscaleWithModel