sniklaus/softmax-splatting
A reference PyTorch implementation of the softmax splatting operator for differentiable forward warping in video frame interpolation.
Too new to rank. Velocity needs a few days of history.
star history
This repository implements softmax splatting, a differentiable forward warping operator proposed in the paper ‘Softmax Splatting for Video Frame Interpolation’. The operator is written in CUDA via CuPy and used from PyTorch. Included scripts run interpolation on image pairs or videos and benchmark results on Middlebury optical flow examples.
Frequently asked
- What is sniklaus/softmax-splatting?
- A reference PyTorch implementation of the softmax splatting operator for differentiable forward warping in video frame interpolation.
- Is softmax-splatting open source?
- Yes — sniklaus/softmax-splatting is an open-source project tracked on heatdrop.
- What language is softmax-splatting written in?
- sniklaus/softmax-splatting is primarily written in Python.
- How popular is softmax-splatting?
- sniklaus/softmax-splatting has 512 stars on GitHub.
- Where can I find softmax-splatting?
- sniklaus/softmax-splatting is on GitHub at https://github.com/sniklaus/softmax-splatting.