[ICLR 2026] D²GS: Depth-and-Density Guided Gaussian Splatting for Stable and Accurate Sparse-View Reconstruction

Meixi Song* · Xin Lin* · Dizhe Zhang · Haodong Li
Xiangtai Li · Bo Du · Lu Qi

Project Page

Overview

The proposed D²GS mainly consists of two key components: a Depth-and-Density guided Dropout (DD-Drop) mechanism and Distance-Aware Fidelity Enhancement (DAFE), to improve the stability and spatial completeness of scene reconstruction under sparse-view settings. DD-Drop assigns each Gaussian a dropout score based on local density and camera distance, indicating regions prone to overfitting. High-scoring Gaussians would be dropped with a higher probability to suppress aliasing and improve rendering fidelity. In addition, DAFE avoids underfitting by boosting supervision in distant regions using depth priors.

Implementation

Installation

We provide an installation using Conda package and environment management:

git clone https://github.com/Willacold/DDGS
cd DDGS
conda env create --file environment.yaml
conda activate DDGS

Note: This Conda environment assumes that CUDA 12.1 is already installed on your system.

Data Preparation

In the data preparation stage, we first reconstruct sparse-view inputs using Structure-from-Motion (SfM) with the provided camera poses from the datasets. Then, we perform dense stereo matching using COLMAP’s patch_match_stereo function, followed by stereo_fusion to generate the dense stereo point cloud.

mkdir dataset
cd dataset

# Download LLFF dataset
gdown 16VnMcF1KJYxN9QId6TClMsZRahHNMW5g

# Generate sparse point cloud using COLMAP (limited views) for LLFF
python tools/colmap_llff.py

# Download MipNeRF-360 dataset
wget http://storage.googleapis.com/gresearch/refraw360/360_v2.zip
unzip -d mipnerf360 360_v2.zip

# Generate sparse point cloud using COLMAP (limited views) for MipNeRF-360
python tools/colmap_360.py

Training

LLFF Dataset

To train on a single LLFF scene, use the following command:

python train.py -s dataset/nerf_llff_data/trex -m output_llff --depth_weight 0.15 --density_weight 0.85 --drop_min 0.05 --drop_max 0.5 --mask_param 5 --lambda_far 0.5 --eval -r 8 --n_views 3

To train and evaluate on all LLFF scenes, simply run the script below:

bash scripts/train_llff.sh

MipNeRF-360 Dataset

To train on a single MipNeRF-360 scene, use the following command:

python train.py -s dataset/mipnerf360/kitchen -m output_mipnerf360 --depth_weight 0.1 --density_weight 0.9 --drop_min 0.05 --drop_max 0.5 --mask_param 10 --lambda_far 0.5 --eval -r 8 --n_views 12

To train and evaluate on all MipNeRF-360 scenes, simply run the script below:

bash scripts/train_mipnerf360.sh

Rendering & Evaluation

You can perform rendering and evaluation in a single step using the following command:

LLFF Dataset

python render.py -s -m output_llff --eval -r 8

MipNeRF-360 Dataset

python render.py -s -m output_mipnerf360 --eval -r 8

Acknowledgement

Our implementation and experiments are built on top of open-source GitHub repositories. We thank all the authors who made their code public, which tremendously accelerates our project progress. If you find these works helpful, please consider citing them as well.

DCVL-3D/DropGaussian_release
GraphDeco-INRIA/gaussian-splatting