Back to workflows

Multi-Keyframe Video Stitching

Creates a smooth video using 6 key frames.It auto fills in the motion between frames and stitchs the segments together seamlessly.

Download WorkflowTry in Cloud

Input

The first frame of the video.

Segment 1 ends and segment 2 starts here.

Segment 2 ends and segment 3 starts here.

Segment 3 ends and segment 4 starts here.

Segment 4 ends and segment 5 starts here.

The last frame of the video.

How the scene moves from keyframe 1 to keyframe 2.

How the scene moves from keyframe 2 to keyframe 3.

How the scene moves from keyframe 3 to keyframe 4.

How the scene moves from keyframe 4 to keyframe 5.

How the scene moves from keyframe 5 to keyframe 6.

Every segment is generated at this size; 832×480 for widescreen.

Frames between two keyframes; 25 is about 1 s. Use 4n+1 values.

Every segment is generated at this size; 832×480 for widescreen.

Advanced settings

Frame rate of the stitched video.

Default: 24

What to avoid, used for every segment.

Leave empty for a new result each run. Enter a number to repeat one.

Run

Output
00:00
Example

Try an example

About this workflow

The Multi-Keyframe Video Stitching workflow in ComfyUI is designed to transform a series of static images into a fluid video sequence. Utilizing six key frames, this workflow automatically generates the transitional frames needed to create a seamless video. At the core of this process is the Wan2.2 model, which excels at interpreting motion between images to ensure smooth transitions.

The workflow begins with the LoadImage and ImageBatch nodes, which prepare the key frames for processing. The intermediate nodes, such as 11ac0cb0-38fb-41d4-9a55-7c44422b2516 and 1c5f4533-5d1e-4449-97ca-9027b9ea9c11, handle the interpolation of frames by predicting and filling in the motion paths. Finally, the CreateVideo and SaveVideo nodes compile the frames into a cohesive video file.

This workflow is particularly useful for creators looking to produce animations or video content from still images without requiring extensive manual editing. By automating the interpolation process, it saves time and ensures consistency in the motion between frames. The use of advanced nodes and the Wan2.2 model allows for high-quality video output that maintains the integrity of the original images while providing a natural flow of motion.

Questions & answers