OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

3D Ken Burns

3D Ken Burns is a Python/PyTorch reference implementation that animates one still image with a virtual camera scan and zoom, using motion parallax to create a 3D Ken Burns effect.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

3D Ken Burns — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when a still image needs simulated camera movement and motion parallax, especially when an automatic result or a manually adjusted camera path suits the workflow.

Who it is for

  • Developers and researchers who want to generate parallax camera-movement clips from still images.
  • Creators seeking animated pan-and-zoom previews for photo galleries or still-image presentations.

Documented uses

  • Generate an automatic camera scan and zoom video from one image.
  • Manually adjust the camera path through a local browser interface.
  • Produce a raw depth estimate from an image.
Before you choose

Requirements & limits

  • The README requires CuPy and a configured CUDA_HOME because several functions use CUDA.
  • MoviePy is required for generating video results.
  • The manual adjustment workflow runs a local interface at localhost:8080.
  • The depth-estimation script produces a raw estimate and does not perform depth adjustment.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Generate an automatic parallax video from an image

Use the documented automatic workflow to create a 3D Ken Burns video file from one still image.

Before you start

  • Python project environment with the repository available.
  • CuPy installed and CUDA_HOME configured; several functions use CUDA via CuPy.
  • MoviePy installed to generate video results.
  • An input image.

Steps

  1. Install CuPy with the documented pip command or use an appropriate CuPy binary package, and configure CUDA_HOME.

  2. Install MoviePy.

  3. Run the command below, replacing the example input and output paths as needed.

    Documented command
    python autozoom.py --in ./images/doublestrike.jpg --out ./autozoom.mp4

Expected outcome

The command is documented to generate an automatically animated 3D Ken Burns video at the specified output path.

Guide limitations

  • Several functions use CUDA through CuPy, and the README requires CUDA_HOME to be configured.
  • The documented automatic command provides no camera-path adjustment; use the interface workflow for manual camera-path control.
Documentation 1 ↗
Source-based practical guide

Open the local interface to adjust a camera path

Start the documented interface and load an image for manual camera-path adjustment.

Before you start

  • Python project environment with the repository available.
  • CuPy installed and CUDA_HOME configured.
  • MoviePy installed to generate video results.

Steps

  1. Run the command below.

    Documented command
    python interface.py

  2. Open http://localhost:8080/ in a browser.

  3. Load an image using the button at the bottom right, then adjust the camera path and save the result.

Expected outcome

The local interface provides a workflow for manually adjusting the camera path for an image.

Guide limitations

  • The README notes that image loading and saving involve background processing and may take patience.
  • Several functions use CUDA through CuPy, and video generation requires MoviePy.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
17 October 2026
Original website title
sniklaus/3d-ken-burns · Official homepage title · retained in its source language
How we review sourcesSuggest a correction