Figment Nodes

This is a list of all the nodes in Figment:

Core

• Null: Does nothing.
• Out: Signifies that this is the output of the network.
• Shortcut Trigger: Turn a global keyboard shortcut into a boolean signal.

Image Operations

Creating / Loading Images

• Constant: Render a constant color.
• Drawing: Draw by hand in a browser window.
• Fetch Image: Fetch an image from the internet.
• Load Image: Load an image from a file.
• Load Image Folder: Load a folder of images.
• Load Movie: Load a movie file.
• Webcam Image: Return a webcam stream.

Saving Images

• Save Image: Save the image sequence to disk during export.

Resizing / Cropping / Combining images

• Resize: Resize the input image.
• Crop: Crop an input image.
• Composite: Combine two images together.
• Conditional: Pick one of two images based on a boolean.
• Stack: Combine 2 images horizontally / vertically.
• Mask Image: Mask the input image with another image.
• Mask Ellipse: Draw a circular mask of an image or color.
• Color Key: Make pixels of a certain color transparent.
• Projection Quad: Map the image onto a four-cornered surface for projection mapping.

Filters / Effects

The images of the examples courtesy of John Mark Arnold and Sergey Shmidt.

Color and tone:

• Bleach Bypass: High-contrast, desaturated film look.
• Brannan: Brannan filter.
• Center Around Gray: Keep the color inside a circle, grayscale outside.
• Colorify: Repaint the image in one color.
• Grayscale: Convert the input image to grayscale.
• Instagram Filters: A set of Instagram-like filters.
• Invert: Invert the colors of input image.
• Levels: Change the brightness/contrast/saturation of an image.
• Lookup: Map the colors of one image to another image.
• Modulate Color: Adjust the colors of the input image.
• Reduce Color: Reduce the amount of colors of input image.
• Gray Cluster: Cluster the image into a few gray levels.
• Rgb Cluster: Cluster the image into a few colors.
• Sepia: Sepia tone.
• Solarize: Invert the channels above a threshold.
• Technicolor: Two-strip Technicolor look.
• Threshold: Change brightness threshold of input image.
• Vignette: Darken the edges of the image.

Blur, sharpen, and edges:

• Blur: Blur an input image.
• Gaussian Blur: A light, cheap blur.
• Denoise: Remove noise from the image.
• Sharpen: Sharpen an input image
• Canny: Canny edge detection on input image.
• Freichen: Frei-Chen edge detection.
• Glow Edges: Glowing edges.
• INMS: Intensity-based non-maximum suppression edge detection.
• LoG: Laplacian of Gaussian edge detection.
• Sobel: Sobel edge detection on input image.
• Emboss: Emboss convolution on an input image.

Stylize:

• Ascii: Render the image as character glyphs.
• Cartoon: Cartoon look.
• Glitch: Digital glitch effect.
• Heatmap: Color a depth map in five bands.
• Noise: Mix random grain into the image.
• Pixelate: Pixelate the input image.
• Reaction Diffusion: Grow organic patterns seeded by the image.

Geometry and distortion:

• Barrel Distortion: Barrel lens distortion.
• Border: Generate a border around the image.
• Chromatic: Chromatic aberration.
• Distortion: Distort the image.
• Kaleidoscope: Mirror a wedge of the image around its center.
• Lens Distortion: Distort an image using a lens distortion shader.
• Mirror: Mirror the input image over a specific axis.
• Radial Distortion: Radial lens distortion.
• Screen Distortion: Bulge the image like an old monitor.
• Transform: Transform the image.
• Wrap: Bend the image into a circle.

Over time:

• Difference: Show what changed since the previous frame.
• Smooth: Blend each frame with the frames before it.
• Trail: Don’t erase the previous input image, creating a trail.

Machine Learning

• Detect Faces: Detect faces in an image.
• Detect Hands: Detect hands in an input image.
• Detect Pose: Detect human poses in input image.
• Normalize Pose: Scale and move pose landmarks to match the body a model was trained on.
• Normalize Face: Scale and move face landmarks to match the face a model was trained on.
• Draw Landmarks: Draw pose, hand or face landmarks as an image.
• Segment Pose: Cut people out of an image using the pose model.
• Segment Image: Cut the person out of a portrait.
• ONNX Image Model: Run your own image to image model (pix2pix).

Audio

• Audio Spectrum: Play audio and draw its frequency spectrum.
• Audio Waveform: Play audio and draw its waveform.

Communication

• Send OSC: Send OSC messages, including detected landmarks, to other applications.
• Receive Rokoko: Receive live motion capture from Rokoko Studio.
• Share Image: Share the image with other applications via Syphon (macOS).