Skip to main content

3DGS Floater Removal — Auto-Clean Your Gaussian Splatting Scans

Automatically detect and remove floating artifacts from 3DGS scenes using statistical outlier analysis. Runs in your browser.

Last updated Mar 2026

Drop a 3DGS file to remove floaters

Supports .ply and .splat (max 500 MB)

What You Should Know

What Are Floaters in Gaussian Splatting?

Floaters are stray Gaussians that appear as floating blobs or specks scattered around a 3DGS scene. They are caused by reconstruction errors during the training process — areas with insufficient camera coverage, reflective surfaces, or moving objects produce Gaussians that are not attached to any real surface. Floaters degrade visual quality, increase file size, and consume GPU memory during rendering. Removing them is one of the most common post-processing steps after scanning.

How Automatic Detection Works

The tool uses Statistical Outlier Removal (SOR), a well-established algorithm from point cloud processing. For each Gaussian, it computes the average distance to its K nearest neighbors using a K-D tree spatial index. Gaussians that are significantly farther from their neighbors than the global average (beyond a threshold of mean + N standard deviations) are flagged as outliers. Additionally, Gaussians with very low opacity (nearly invisible) are flagged since they typically represent reconstruction noise rather than meaningful scene content.

Adjusting the Sensitivity

The sigma (σ) slider controls how aggressively floaters are detected. A lower value (1.0-1.5σ) removes more Gaussians including borderline cases — good for heavily noisy scans but may clip thin structures like antennas or hair. A higher value (2.5-3.5σ) is conservative, removing only clearly isolated points. The default of 2.0σ works well for most indoor and outdoor scans. The K (neighbors) slider affects accuracy: more neighbors give smoother statistical estimates but take longer to compute.

Performance and Limitations

The tool processes files up to 500 MB (~2M Gaussians for PLY with SH degree 0). Processing time depends on Gaussian count and the K parameter: 100K Gaussians with K=30 takes about 10-20 seconds, while 500K Gaussians may take 1-2 minutes. The K-D tree construction and nearest-neighbor queries run on the main thread, so the browser tab may be unresponsive during processing. For files larger than 500 MB, consider using desktop tools like SuperSplat or the Open3D Python library.

Frequently Asked Questions
SuperSplat requires you to visually identify floaters and manually select them with brush or box tools. This tool automatically detects outliers using statistical analysis — no visual inspection needed. It finds floaters you might miss, especially small scattered ones far from the main scene. For complex cleanup, you can use this tool first for bulk removal, then fine-tune in SuperSplat.
At the default sensitivity (2.0σ), the tool only removes clearly isolated Gaussians. However, thin structures (antennas, wires, hair strands) and sparse edges may be affected because their Gaussians are naturally farther apart. If you see important details being removed, increase the sigma value to be more conservative. Always preview the before/after Gaussian count before downloading.
The tool does not modify your original file. It creates a new cleaned file for download. Your original file remains unchanged on your device. If the result is too aggressive, adjust the sigma and K parameters and run again.
Yes. SOR works on any point distribution. Indoor scans typically have fewer floaters (controlled lighting), while outdoor scans (sky, reflections, vegetation) tend to have more. Outdoor scans may need a slightly lower sigma (1.5-2.0) for effective cleanup.
No. All processing happens locally in your browser. The K-D tree is built in memory, outliers are detected, and the cleaned file is generated — nothing leaves your device.
The tool handles scenes up to approximately 2 million Gaussians. Memory usage is several times the input file size (positions array + K-D tree nodes + output buffer). For larger scenes, use desktop tools like SuperSplat or Open3D.

Related Tools

Step-by-Step Guides