DMD³C: Distilling Monocular Foundation Models for Fine-grained Depth Completion

Official implementation of the CVPR 2025 paper "Distilling Monocular Foundation Models for Fine-grained Depth Completion"

---

_{Url: KITTI Depth Completion Benchmark}

---

🔍 Overview

Depth completion methods often suffer in regions with sparse or missing supervision, leading to inaccurate fine-grained structures and degraded depth quality.

DMD³C introduces a novel framework that distills rich geometric priors from monocular foundation models into the depth completion pipeline. By leveraging dense knowledge from foundation models, DMD³C significantly improves depth estimation quality, particularly in regions lacking ground-truth supervision.

Key Features

• Distills geometric knowledge from monocular foundation models
• Enhances fine-grained structure recovery
• Improves depth estimation in sparse and unsupervised regions
• Achieves strong performance on benchmark datasets

---

---

🚀 Getting Started

0. Conda Environment

You can directly build the environment by running the following command if you use conda as the environment management tool.

conda env create -f environment.yml

Compile the C++ and CUDA code:

cd exts
python setup.py install

1. Dataset Preparation

Please follow the dataset preparation instructions from:

👉 BP-Net

The structure of data directory should be:

└── datas
└── kitti
├── data_depth_annotated
│ ├── train
│ └── val
├── data_depth_velodyne
│ ├── train
│ └── val
├── raw
│ ├── 2011_09_26
│ ├── 2011_09_28
│ ├── 2011_09_29
│ ├── 2011_09_30
│ └── 2011_10_03
├── test_depth_completion_anonymous
│ ├── image
│ ├── intrinsics
│ └── velodyne_raw
└── val_selection_cropped
├── groundtruth_depth
├── image
├── intrinsics
└── velodyne_raw

2. Training

Run the training script:

bash train.sh

Our models are trained on 8 GPU workstation with Nvidia GTX 4090 (48G).

3. Pretrained Models

Download pretrained checkpoints from:

👉 Hugging Face

Place the .pth file into "./checkpoints/PMP_Residual_Norm_ssil_KITTI/"

4. Submission

Generate predictions and submit results to the KITTI online benchmark:

bash submission.sh

The results will be save into "./results" folder.

---

---

🌍 DCVerse Benchmark

Download

To facilitate fair and reproducible evaluation of depth completion methods, we build DCVerse, a unified depth completion benchmark that standardizes the experimental settings across different methods and datasets.

DCVerse addresses inconsistencies commonly found in previous evaluations, including:

• Unified input image resolution
• Consistent sparse point sampling density
• Standardized sparse sampling strategies
• Unified evaluation metrics
• Cross-dataset evaluation protocol

The benchmark enables more reliable and direct comparisons between different depth completion methods.

The benchmark and processed data can be found at:

👉 DCVerse on Hugging Face

---

Usage

#!/bin/bash
datasets=(
ETH3D_SfM_Indoor_test
ETH3D_SfM_Outdoor_test
VKITTI2_clone
VKITTI2_fog
VKITTI2_morning
VKITTI2_overcast
VKITTI2_rain
VKITTI2_sunset
KITTIDC_test_LiDAR_64
KITTIDC_test_LiDAR_32
KITTIDC_test_LiDAR_16
KITTIDC_test_LiDAR_8
VOID_sample1500
VOID_sample500
VOID_sample150
NYU_test_500
NYU_test_200
NYU_test_100
NYU_test_50
iBims_test_LiDAR_32
ARKitScenes_test_300
DIODE_Indoor_test_300
DDAD_val
)
mkdir -p results
fordatasetin"${datasets[@]}"doecho"======================================"echo"Running dataset: ${dataset}"echo"======================================"
python test.py \
gpus=[0] \
name=PMP_Residual_Norm_ssil_KITTI_${dataset} \
++chpt=PMP_Residual_Norm_ssil_KITTI \
net=PMP_Residual_Norm_fast \
num_workers=4 \
data=UNI \
data.testset.mode=test \
data.path=/PATH-TO-DATA/${dataset} \
test_batch_size=1 \
metric=MetricALL \
++save=true \
2>&1| tee "results/${dataset}.log"doneecho"All tests finished."

The adapted implementations are available in the benchmarks/ directory.

Supported Methods

The current benchmark includes implementations of the following representative depth completion methods:

Category	Methods
Classical Depth Completion	LRRU, VPP4DC, CompletionFormer, ImprovingDC, BP-Net, DepthPrompting, OGNI-DC, DMD3C
Zero-shot Models	G2-MD, Marigold-DC, SPNet, OMNI-DC, PacGDC
Ours	Coming Soon

We continuously maintain and extend the benchmark to include newly proposed methods and stronger baselines, providing a unified platform for fair and reproducible depth completion evaluation.

📝 Citation

If you find our work useful for your research, please consider citing:

@inproceedings{liang2025distilling,
title={Distilling Monocular Foundation Models for Fine-grained Depth Completion},
author={Liang, Yingping and Hu, Yutao and Shao, Wenqi and Fu, Ying},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
pages={22254--22265},
year={2025}
}

---

🙏 Acknowledgement

This project is built upon and inspired by:

• BP-Net

• OMNI-DC

We sincerely thank the authors for making their code publicly available.