naver/mast3r
naver/mast3r sits at 3.1k stars on GitHub, written primarily in Python. Grounding Image Matching in 3D with MASt3R
Snapshot summary built from the project's own GitHub metadata — there's no written TopGit review yet. The page will update automatically when a full review is published.
TopGit writes full reviews for the most-starred, most-requested repositories. This page is a snapshot until then — see the READ ME tab for the original README in full.
Snapshot
Top contributors
Show top contributors
Official implementation of Grounding Image Matching in 3D with MASt3R
[Project page], [MASt3R arxiv], [DUSt3R arxiv]
@misc{mast3r_eccv24,
title={Grounding Image Matching in 3D with MASt3R},
author={Vincent Leroy and Yohann Cabon and Jerome Revaud},
booktitle = {ECCV},
year = {2024}
}
@misc{mast3r_arxiv24,
title={Grounding Image Matching in 3D with MASt3R},
author={Vincent Leroy and Yohann Cabon and Jerome Revaud},
year={2024},
eprint={2406.09756},
archivePrefix={arXiv},
primaryClass={cs.CV}
}
@inproceedings{dust3r_cvpr24,
title={DUSt3R: Geometric 3D Vision Made Easy},
author={Shuzhe Wang and Vincent Leroy and Yohann Cabon and Boris Chidlovskii and Jerome Revaud},
booktitle = {CVPR},
year = {2024}
}
@inproceedings{
duisterhof2025mastrsfm,
title={{MAS}t3R-SfM: a Fully-Integrated Solution for Unconstrained Structure-from-Motion},
author={Bardienus Pieter Duisterhof and Lojze Zust and Philippe Weinzaepfel and Vincent Leroy and Yohann Cabon and Jerome Revaud},
booktitle={International Conference on 3D Vision 2025},
year={2025},
url={https://openreview.net/forum?id=5uw1GRBFoT}
}
Table of Contents
- Table of Contents
- License
- Get Started
- Installation
- Checkpoints
- MASt3R Model
- Retrieval Model
- Dune Model
- MASt3R-SfM
- Interactive demo
- Interactive demo with docker
- Usage
- Usage MASt3R
- Usage DUNE+MASt3R
- Training
- Datasets
- Demo
- Our Hyperparameters
- Visual Localization
- Dataset preparation
- Example Commands
License
The code is distributed under the CC BY-NC-SA 4.0 License. See LICENSE for more information.
# Copyright (C) 2024-present Naver Corporation. All rights reserved.
# Licensed under CC BY-NC-SA 4.0 (non-commercial use only).
Get Started
Installation
- Clone MASt3R.
git clone --recursive https://github.com/naver/mast3r
cd mast3r
# if you have already cloned mast3r:
# git submodule update --init --recursive
- Create the environment, here we show an example using conda.
conda create -n mast3r python=3.11 cmake=3.14.0
conda activate mast3r
conda install pytorch torchvision pytorch-cuda=12.1 -c pytorch -c nvidia # use the correct version of cuda for your system
pip install -r requirements.txt
pip install -r dust3r/requirements.txt
# Optional: you can also install additional packages to:
# - add support for HEIC images
# - add required packages for visloc.py
pip install -r dust3r/requirements_optional.txt
- compile and install ASMK
pip install cython
git clone https://github.com/jenicek/asmk
cd asmk/cython/
cythonize *.pyx
cd ..
pip install . # or python3 setup.py build_ext --inplace
cd ..
- Optional, compile the cuda kernels for RoPE (as in CroCo v2).
# DUST3R relies on RoPE positional embeddings for which you can compile some cuda kernels for faster runtime.
cd dust3r/croco/models/curope/
python setup.py build_ext --inplace
cd ../../../../
Checkpoints
MASt3R Model
You can obtain the model checkpoints by two ways:
-
You can use our huggingface_hub integration: the models will be downloaded automatically.
-
Otherwise, download it from our server:
| Modelname | Training resolutions | Head | Encoder | Decoder |
|---|---|---|---|---|
MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric | 512x384, 512x336, 512x288, 512x256, 512x160 | CatMLP+DPT | ViT-L | ViT-B |
You can check the hyperparameters we used to train these models in the section: Our Hyperparameters Make sure to check license of the datasets we used.
To download MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric.pth:
mkdir -p checkpoints/
wget https://download.europe.naverlabs.com/ComputerVision/MASt3R/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric.pth -P checkpoints/
Make sure to agree to the license of all the training datasets we used, in addition to CC-BY-NC-SA 4.0. The mapfree dataset license in particular is very restrictive. For more information, check CHECKPOINTS_NOTICE.
Retrieval Model
This retrieval model is for MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric only.
You need to download both the trainingfree.pth and codebook.pkl files, and put them in the same directory.
MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric_retrieval_trainingfree
MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric_retrieval_codebook
mkdir -p checkpoints/
wget https://download.europe.naverlabs.com/ComputerVision/MASt3R/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric_retrieval_trainingfree.pth -P checkpoints/
wget https://download.europe.naverlabs.com/ComputerVision/MASt3R/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric_retrieval_codebook.pkl -P checkpoints/
Dune Model
We added partial support of the Dune encoder. Check the associated Dune License.
You can find the MASt3R decoder that goes with it here:
dunemast3r_cvpr25_vitbase
dunemast3r_cvpr25_vitsmall
mkdir -p checkpoints/
wget https://download.europe.naverlabs.com/dune/dunemast3r_cvpr25_vitbase.pth -P checkpoints/
This model have limited compatility with the rest of the codebase, but we wanted to include it as it achieves impressive results on the Map-free Visual Relocalization benchmark. Make sure to check the Usage DUNE+MASt3R section if you are interested.
MASt3R-SfM
A few words about the addition of MASt3R-SfM to this repository.
MASt3R-SfM refers to the make_pairs (retrieval) + sparse_global_alignment that you can find here: demo.py#L142.
In this repository, you will also find kapture_mast3r_mapping.py and demo_glomap.py. These two scripts are unrelated to MASt3R-SfM. They are "toys" that attempt to use mast3r matches to do standard Sfm reconstructions with colmap/glomap. As such, they were not extensively tested, and may fail on edge cases.
Interactive demo
We made one huggingface space running the new sparse global alignment in a simplified demo for small scenes: naver/MASt3R There are two demos available to run locally:
demo.py is the updated demo for MASt3R. It uses our new sparse global alignment method that allows you to reconstruct larger scenes
python3 demo.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric
# Use --weights to load a checkpoint from a local file, eg --weights checkpoints/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric.pth
# Use --retrieval_model and point to the retrieval checkpoint (*trainingfree.pth) to enable retrieval as a pairing strategy, asmk must be installed
# Use --local_network to make it accessible on the local network, or --server_name to specify the url manually
# Use --server_port to change the port, by default it will search for an available port starting at 7860
# Use --device to use a different device, by default it's "cuda"
demo_dust3r_ga.py is the same demo as in dust3r (+ compatibility for MASt3R models)
see https://github.com/naver/dust3r?tab=readme-ov-file#interactive-demo for details
Interactive demo with docker
TODO update with asmk/retrieval model
To run MASt3R using Docker, including with NVIDIA CUDA support, follow these instructions:
-
Install Docker: If not already installed, download and install
dockeranddocker composefrom the Docker website. -
Install NVIDIA Docker Toolkit: For GPU support, install the NVIDIA Docker toolkit from the Nvidia website.
-
Build the Docker image and run it:
cdinto the./dockerdirectory and run the following commands:
cd docker
bash run.sh --with-cuda --model_name="MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric"
Or if you want to run the demo without CUDA support, run the following command:
cd docker
bash run.sh --model_name="MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric"
By default, demo.py is launched with the option --local_network.
Visit http://localhost:7860/ to access the web UI (or replace localhost with the machine's name to access it from the network).
run.sh will launch docker-compose using either the docker-compose-cuda.yml or docker-compose-cpu.ym config file, then it starts the demo using entrypoint.sh.
Usage
Usage MASt3R
Code sample to compute matches with MASt3R for a pair of images
from mast3r.model import AsymmetricMASt3R
from mast3r.fast_nn import fast_reciprocal_NNs
import mast3r.utils.path_to_dust3r
from dust3r.inference import inference
from dust3r.utils.image import load_images
if __name__ == '__main__':
device = 'cuda'
model_name = "naver/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric"
# you can put the path to a local checkpoint in model_name if needed
model = AsymmetricMASt3R.from_pretrained(model_name).to(device)
images = load_images(['dust3r/croco/assets/Chateau1.png', 'dust3r/croco/assets/Chateau2.png'], size=512)
output = inference([tuple(images)], model, device, batch_size=1, verbose=False)
# at this stage, you have the raw dust3r predictions
view1, pred1 = output['view1'], output['pred1']
view2, pred2 = output['view2'], output['pred2']
desc1, desc2 = pred1['desc'].squeeze(0).detach(), pred2['desc'].squeeze(0).detach()
# find 2D-2D matches between the two images
matches_im0, matches_im1 = fast_reciprocal_NNs(desc1, desc2, subsample_or_initxy1=8,
device=device, dist='dot', block_size=2**13)
# ignore small border around the edge
H0, W0 = view1['true_shape'][0]
valid_matches_im0 = (matches_im0[:, 0] >= 3) & (matches_im0[:, 0] < int(W0) - 3) & (
matches_im0[:, 1] >= 3) & (matches_im0[:, 1] < int(H0) - 3)
H1, W1 = view2['true_shape'][0]
valid_matches_im1 = (matches_im1[:, 0] >= 3) & (matches_im1[:, 0] < int(W1) - 3) & (
matches_im1[:, 1] >= 3) & (matches_im1[:, 1] < int(H1) - 3)
valid_matches = valid_matches_im0 & valid_matches_im1
matches_im0, matches_im1 = matches_im0[valid_matches], matches_im1[valid_matches]
# visualize a few matches
import numpy as np
import torch
import torchvision.transforms.functional
from matplotlib import pyplot as pl
n_viz = 20
num_matches = matches_im0.shape[0]
match_idx_to_viz = np.round(np.linspace(0, num_matches - 1, n_viz)).astype(int)
viz_matches_im0, viz_matches_im1 = matches_im0[match_idx_to_viz], matches_im1[match_idx_to_viz]
image_mean = torch.as_tensor([0.5, 0.5, 0.5], device='cpu').reshape(1, 3, 1, 1)
image_std = torch.as_tensor([0.5, 0.5, 0.5], device='cpu').reshape(1, 3, 1, 1)
viz_imgs = []
for i, view in enumerate([view1, view2]):
rgb_tensor = view['img'] * image_std + image_mean
viz_imgs.append(rgb_tensor.squeeze(0).permute(1, 2, 0).cpu().numpy())
H0, W0, H1, W1 = *viz_imgs[0].shape[:2], *viz_imgs[1].shape[:2]
img0 = np.pad(viz_imgs[0], ((0, max(H1 - H0, 0)), (0, 0), (0, 0)), 'constant', constant_values=0)
img1 = np.pad(viz_imgs[1], ((0, max(H0 - H1, 0)), (0, 0), (0, 0)), 'constant', constant_values=0)
img = np.concatenate((img0, img1), axis=1)
pl.figure()
pl.imshow(img)
cmap = pl.get_cmap('jet')
for i in range(n_viz):
(x0, y0), (x1, y1) = viz_matches_im0[i].T, viz_matches_im1[i].T
pl.plot([x0, x1 + W0], [y0, y1], '-+', color=cmap(i / (n_viz - 1)), scalex=False, scaley=False)
pl.show(block=True)
Usage DUNE+MASt3R
At the moment, you can only do two things:
- Extract matches, following the subset of code below
- Run the
demo_dust3r_ga.pyscript with option--weights checkpoints/dunemast3r_cvpr25_vitbase.pth --image_size 518
Code sample to compute matches with DUNE+MASt3R for a pair of images
from mast3r.model import load_dune_mast3r_model
from mast3r.fast_nn import fast_reciprocal_NNs
import mast3r.utils.path_to_dust3r # noqa
from dust3r.utils.image import load_images
from dust3r.inference import inference
import torch
if __name__ == '__main__':
device = torch.device('cuda:0')
model = load_dune_mast3r_model('checkpoints/dunemast3r_cvpr25_vitbase.pth', device)
images = load_images(['dust3r/croco/assets/Chateau1.png', 'dust3r/croco/assets/Chateau2.png'],
size=518, patch_size=model.patch_size, square_ok=True)
output = inference([tuple(images)], model, device, batch_size=1, verbose=False)
# at this stage, you have the raw dust3r predictions
view1, pred1 = output['view1'], output['pred1']
view2, pred2 = output['view2'], output['pred2']
desc1, desc2 = pred1['desc'].squeeze(0).detach(), pred2['desc'].squeeze(0).detach()
# find 2D-2D matches between the two images
matches_im0, matches_im1 = fast_reciprocal_NNs(desc1, desc2, subsample_or_initxy1=8,
device=device, dist='dot', block_size=2**13)
# ignore small border around the edge
H0, W0 = view1['true_shape'][0]
valid_matches_im0 = (matches_im0[:, 0] >= 3) & (matches_im0[:, 0] < int(W0) - 3) & (
matches_im0[:, 1] >= 3) & (matches_im0[:, 1] < int(H0) - 3)
H1, W1 = view2['true_shape'][0]
valid_matches_im1 = (matches_im1[:, 0] >= 3) & (matches_im1[:, 0] < int(W1) - 3) & (
matches_im1[:, 1] >= 3) & (matches_im1[:, 1] < int(H1) - 3)
valid_matches = valid_matches_im0 & valid_matches_im1
matches_im0, matches_im1 = matches_im0[valid_matches], matches_im1[valid_matches]
# visualize a few matches
import numpy as np
import torch
import torchvision.transforms.functional
from matplotlib import pyplot as pl
n_viz = 20
num_matches = matches_im0.shape[0]
match_idx_to_viz = np.round(np.linspace(0, num_matches - 1, n_viz)).astype(int)
viz_matches_im0, viz_matches_im1 = matches_im0[match_idx_to_viz], matches_im1[match_idx_to_viz]
image_mean = torch.as_tensor([0.5, 0.5, 0.5], device='cpu').reshape(1, 3, 1, 1)
image_std = torch.as_tensor([0.5, 0.5, 0.5], device='cpu').reshape(1, 3, 1, 1)
viz_imgs = []
for i, view in enumerate([view1, view2]):
rgb_tensor = view['img'] * image_std + image_mean
viz_imgs.append(rgb_tensor.squeeze(0).permute(1, 2, 0).cpu().numpy())
H0, W0, H1, W1 = *viz_imgs[0].shape[:2], *viz_imgs[1].shape[:2]
img0 = np.pad(viz_imgs[0], ((0, max(H1 - H0, 0)), (0, 0), (0, 0)), 'constant', constant_values=0)
img1 = np.pad(viz_imgs[1], ((0, max(H0 - H1, 0)), (0, 0), (0, 0)), 'constant', constant_values=0)
img = np.concatenate((img0, img1), axis=1)
pl.figure()
pl.imshow(img)
cmap = pl.get_cmap('jet')
for i in range(n_viz):
(x0, y0), (x1, y1) = viz_matches_im0[i].T, viz_matches_im1[i].T
pl.plot([x0, x1 + W0], [y0, y1], '-+', color=cmap(i / (n_viz - 1)), scalex=False, scaley=False)
pl.show(block=True)
Training
In this section, we present a short demonstration to get started with training MASt3R.
Datasets
See Datasets section in DUSt3R
Demo
Like for the DUSt3R training demo, we're going to download and prepare the same subset of CO3Dv2 - Creative Commons Attribution-NonCommercial 4.0 International and launch the training code on it. It is the exact same process as DUSt3R. The demo model will be trained for a few epochs on a very small dataset. It will not be very good.
# download and prepare the co3d subset
mkdir -p data/co3d_subset
cd data/co3d_subset
git clone https://github.com/facebookresearch/co3d
cd co3d
python3 ./co3d/download_dataset.py --download_folder ../ --single_sequence_subset
rm ../*.zip
cd ../../..
python3 datasets_preprocess/preprocess_co3d.py --co3d_dir data/co3d_subset --output_dir data/co3d_subset_processed --single_sequence_subset
# download the pretrained dust3r checkpoint
mkdir -p checkpoints/
wget https://download.europe.naverlabs.com/ComputerVision/DUSt3R/DUSt3R_ViTLarge_BaseDecoder_512_dpt.pth -P checkpoints/
# for this example we'll do fewer epochs, for the actual hyperparameters we used in the paper, see the next section: "Our Hyperparameters"
torchrun --nproc_per_node=4 train.py \
--train_dataset "1000 @ Co3d(split='train', ROOT='data/co3d_subset_processed', aug_crop='auto', aug_monocular=0.005, aug_rot90='diff', mask_bg='rand', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], n_corres=8192, nneg=0.5, transform=ColorJitter)" \
--test_dataset "100 @ Co3d(split='test', ROOT='data/co3d_subset_processed', resolution=(512,384), n_corres=1024, seed=777)" \
--model "AsymmetricMASt3R(pos_embed='RoPE100', patch_embed_cls='ManyAR_PatchEmbed', img_size=(512, 512), head_type='catmlp+dpt', output_mode='pts3d+desc24', depth_mode=('exp', -inf, inf), conf_mode=('exp', 1, inf), enc_embed_dim=1024, enc_depth=24, enc_num_heads=16, dec_embed_dim=768, dec_depth=12, dec_num_heads=12, two_confs=True)" \
--train_criterion "ConfLoss(Regr3D(L21, norm_mode='?avg_dis'), alpha=0.2) + 0.075*ConfMatchingLoss(MatchingLoss(InfoNCE(mode='proper', temperature=0.05), negatives_padding=0, blocksize=8192), alpha=10.0, confmode='mean')" \
--test_criterion "Regr3D_ScaleShiftInv(L21, norm_mode='?avg_dis', gt_scale=True, sky_loss_value=0) + -1.*MatchingLoss(APLoss(nq='torch', fp=torch.float16), negatives_padding=12288)" \
--pretrained "checkpoints/DUSt3R_ViTLarge_BaseDecoder_512_dpt.pth" \
--lr 0.0001 --min_lr 1e-06 --warmup_epochs 1 --epochs 10 --batch_size 4 --accum_iter 4 \
--save_freq 1 --keep_freq 5 --eval_freq 1 --disable_cudnn_benchmark \
--output_dir "checkpoints/mast3r_demo"
Our Hyperparameters
We didn't release all the training datasets, but here are the commands we used for training our models:
# MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric - train mast3r with metric regression and matching loss
# we used cosxl to generate variations of DL3DV: "foggy", "night", "rainy", "snow", "sunny" but we were not convinced by it.
torchrun --nproc_per_node=8 train.py \
--train_dataset "57_000 @ Habitat512(1_000_000, split='train', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 68_400 @ BlendedMVS(split='train', mask_sky=True, resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 68_400 @ MegaDepth(split='train', mask_sky=True, resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 45_600 @ ARKitScenes(split='train', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 22_800 @ Co3d(split='train', mask_bg='rand', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 22_800 @ StaticThings3D(mask_bg='rand', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 45_600 @ ScanNetpp(split='train', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 45_600 @ TartanAir(pairs_subset='', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 4_560 @ UnrealStereo4K(resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 1_140 @ VirtualKitti(optical_center_is_centered=True, resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 22_800 @ WildRgbd(split='train', mask_bg='rand', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 145_920 @ NianticMapFree(split='train', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 57_000 @ DL3DV(split='nlight', resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 57_000 @ DL3DV(split='not-nlight', cosxl_augmentations=None, resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5) + 34_200 @ InternalUnreleasedDataset(resolution=[(512, 384), (512, 336), (512, 288), (512, 256), (512, 160)], aug_crop='auto', aug_monocular=0.005, transform=ColorJitter, n_corres=8192, nneg=0.5)" \
--test_dataset "Habitat512(1_000, split='val', resolution=(512,384), seed=777, n_corres=1024) + 1_000 @ BlendedMVS(split='val', resolution=(512,384), mask_sky=True, seed=777, n_corres=1024) + 1_000 @ ARKitScenes(split='test', resolution=(512,384), seed=777, n_corres=1024) + 1_000 @ MegaDepth(split='val', mask_sky=True, resolution=(512,336), seed=777, n_corres=1024) + 1_000 @ Co3d(split='test', resolution=(512,384), mask_bg='rand', seed=777, n_corres=1024)" \
--model "AsymmetricMASt3R(pos_embed='RoPE100', patch_embed_cls='ManyAR_PatchEmbed', img_size=(512, 512), head_type='catmlp+dpt', output_mode='pts3d+desc24', depth_mode=('exp', -inf, inf), conf_mode=('exp', 1, inf), enc_embed_dim=1024, enc_depth=24, enc_num_heads=16, dec_embed_dim=768, dec_depth=12, dec_num_heads=12, two_confs=True, desc_conf_mode=('exp', 0, inf))" \
--train_criterion "ConfLoss(Regr3D(L21, norm_mode='?avg_dis'), alpha=0.2, loss_in_log=False) + 0.075*ConfMatchingLoss(MatchingLoss(InfoNCE(mode='proper', temperature=0.05), negatives_padding=0, blocksize=8192), alpha=10.0, confmode='mean')" \
--test_criterion "Regr3D(L21, norm_mode='?avg_dis', gt_scale=True, sky_loss_value=0) + -1.*MatchingLoss(APLoss(nq='torch', fp=torch.float16), negatives_padding=12288)" \
--pretrained "checkpoints/DUSt3R_ViTLarge_BaseDecoder_512_dpt.pth" \
--lr 0.0001 --min_lr 1e-06 --warmup_epochs 8 --epochs 50 --batch_size 4 --accum_iter 2 \
--save_freq 1 --keep_freq 5 --eval_freq 1 --print_freq=10 --disable_cudnn_benchmark \
--output_dir "checkpoints/MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric"
Visual Localization
Dataset preparation
See Visloc section in DUSt3R
Example Commands
With visloc.py you can run our visual localization experiments on Aachen-Day-Night, InLoc, Cambridge Landmarks and 7 Scenes.
# Aachen-Day-Night-v1.1:
# scene in 'day' 'night'
# scene can also be 'all'
python3 visloc.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric --dataset "VislocAachenDayNight('/path/to/prepared/Aachen-Day-Night-v1.1/', subscene='${scene}', pairsfile='fire_top50', topk=20)" --pixel_tol 5 --pnp_mode poselib --reprojection_error_diag_ratio 0.008 --output_dir /path/to/output/Aachen-Day-Night-v1.1/${scene}/loc
# or with coarse to fine:
python3 visloc.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric --dataset "VislocAachenDayNight('/path/to/prepared/Aachen-Day-Night-v1.1/', subscene='${scene}', pairsfile='fire_top50', topk=20)" --pixel_tol 5 --pnp_mode poselib --reprojection_error_diag_ratio 0.008 --output_dir /path/to/output/Aachen-Day-Night-v1.1/${scene}/loc --coarse_to_fine --max_batch_size 48 --c2f_crop_with_homography
# InLoc
python3 visloc.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric --dataset "VislocInLoc('/path/to/prepared/InLoc/', pairsfile='pairs-query-netvlad40-temporal', topk=20)" --pixel_tol 5 --pnp_mode poselib --reprojection_error_diag_ratio 0.008 --output_dir /path/to/output/InLoc/loc
# or with coarse to fine:
python3 visloc.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric --dataset "VislocInLoc('/path/to/prepared/InLoc/', pairsfile='pairs-query-netvlad40-temporal', topk=20)" --pixel_tol 5 --pnp_mode poselib --reprojection_error_diag_ratio 0.008 --output_dir /path/to/output/InLoc/loc --coarse_to_fine --max_image_size 1200 --max_batch_size 48 --c2f_crop_with_homography
# 7-scenes:
# scene in 'chess' 'fire' 'heads' 'office' 'pumpkin' 'redkitchen' 'stairs'
python3 visloc.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric --dataset "VislocSevenScenes('/path/to/prepared/7-scenes/', subscene='${scene}', pairsfile='APGeM-LM18_top20', topk=1)" --pixel_tol 5 --pnp_mode poselib --reprojection_error_diag_ratio 0.008 --output_dir /path/to/output/7-scenes/${scene}/loc
# Cambridge Landmarks:
# scene in 'ShopFacade' 'GreatCourt' 'KingsCollege' 'OldHospital' 'StMarysChurch'
python3 visloc.py --model_name MASt3R_ViTLarge_BaseDecoder_512_catmlpdpt_metric --dataset "VislocCambridgeLandmarks('/path/to/prepared/Cambridge_Landmarks/', subscene='${scene}', pairsfile='APGeM-LM18_top50', topk=20)" --pixel_tol 5 --pnp_mode poselib --reprojection_error_diag_ratio 0.008 --output_dir /path/to/output/Cambridge_Landmarks/${scene}/loc
Related repositories
Master programming by recreating your favorite technologies from scratch.
A curated meta-list of curated lists organized by technology domain. The repository acts as a directory pointing to hundreds of specialized awesome lists covering programming languages, platforms, frameworks, and tooling. All content is community-contributed under the CC0 public domain dedication.
Public APIs is a community-curated GitHub repository listing free, publicly accessible APIs across a wide range of categories, with auth type, HTTPS, and CORS noted for each entry. It's a browsable reference, not a library to install.
freeCodeCamp is a free, self-paced curriculum for learning to code, published as open source at freeCodeCamp/freeCodeCamp. It's a 501(c)(3) nonprofit funded by donor support, structured around six certifications in its Full-Stack Developer Curriculum, each gated by required projects instead of open-book quizzes. The repository also carries beta language certifications for developers, interview-prep resources, and the code that runs the live freecodecamp.org platform.
Quick answers
Does naver/mast3r have a project website?
No homepage URL was recorded for naver/mast3r in TopGit's last sync. The README tab above frequently contains screenshots and demo links, or check the repository description on GitHub.
Does naver/mast3r have any tags?
TopGit's last sync did not record any GitHub topics for naver/mast3r. GitHub topics appear in the right sidebar of a repository page; that's the authoritative place to check.
How active is development on naver/mast3r?
The most recent commit recorded on naver/mast3r was 1.1 years ago, based on the GitHub push timestamp. The repository has 279 forks — one of the better signals of community interest.
Is naver/mast3r open source?
TopGit's metadata for naver/mast3r does not record a license. Most public repositories on GitHub ARE open source, but the exact terms vary — verify by opening the LICENSE file directly.
Where do I read more about naver/mast3r?
This TopGit page is a snapshot — the READ ME tab shows the project's own README content (links stripped, images preserved). The GitHub repository at github.com/naver/mast3r is the definitive source.
Read full README in the tab above.
Want a second opinion on mast3r?
Ask an AI that can read this page — one click and you get its take on mast3r.