sniklaus/3d-ken-burns
PyTorch-based neural network that estimates depth from a single image and generates animated 3D Ken Burns video with virtual camera motion and motion parallax.
Not currently ranked — collecting fresh signals.
star history
This repository provides a reference implementation of depth estimation from a single image using a deep learning model. Given an input image, it estimates a depth map and then animates the still image by generating a virtual camera path with zoom and panning, creating motion parallax effects between foreground and background elements. The model uses CUDA via CuPy for GPU-accelerated inference and produces MP4 video output using moviepy.
Frequently asked
- What is sniklaus/3d-ken-burns?
- PyTorch-based neural network that estimates depth from a single image and generates animated 3D Ken Burns video with virtual camera motion and motion parallax.
- Is 3d-ken-burns open source?
- Yes — sniklaus/3d-ken-burns is an open-source project tracked on heatdrop.
- What language is 3d-ken-burns written in?
- sniklaus/3d-ken-burns is primarily written in Python.
- How popular is 3d-ken-burns?
- sniklaus/3d-ken-burns has 1.6k stars on GitHub.
- Where can I find 3d-ken-burns?
- sniklaus/3d-ken-burns is on GitHub at https://github.com/sniklaus/3d-ken-burns.