● Covered by 1 source · 1 reportMedium impact1 neutral
New Method Uses 2D Gaussian Splatting for Bézier Spline Line Art Vectorization
Key points
- New method for line art vectorization using 2D Gaussian splatting.
- Leverages depth prediction and semantic features for stroke extraction.
- Models strokes as Bézier curves with geometric and appearance components.
- Enables efficient fitting and joint optimization of control points and brush texture.
Advancing Line Art Vectorization
A team of researchers from DisneyResearch|Studios, Walt Disney Animation Studios, and ETH Zurich has developed a new technique for vectorizing line art and sketches. This method addresses the challenge of converting raster images into editable vector graphics, a critical task in computer graphics.
Improved Stroke Extraction
The new approach moves beyond traditional heuristic methods for extracting strokes. It incorporates depth prediction and semantic feature extraction models to create a more structured way of splitting the skeleton graph of a sketch into subgraphs. This results in an initial set of strokes that aligns more closely with artistic intent.
Bézier Curves and Gaussian Splatting
The method models individual strokes as Bézier curves, which include both geometric and appearance characteristics. It then employs 2D Gaussian splatting for fast and differentiable rendering. This formulation allows for efficient fitting of strokes to the input image, simultaneously optimizing both the control points of the Bézier curves and the associated brush texture.
Video Integration and Performance
The researchers also demonstrated the method's potential for video applications by integrating temporal tracking and adaptive keyframe introduction. Evaluations indicate that the technique achieves state-of-the-art results for image reconstruction, while maintaining speed and producing high-quality strokes. The fitting process is also designed to be compatible with user input, allowing for corrections to splines and selection of optimization parameters.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Reporting from
Researchers from Disney and ETH Zurich introduced a new method for line art vectorization that uses 2D Gaussian splatting to convert raster images into Bézier splines. This approach improves stroke extraction by leveraging depth prediction and semantic feature models, and enables efficient fitting of strokes with joint optimization of control points and brush texture.