Skip to content

(Arxiv 2021) NeRF--: Neural Radiance Fields Without Known Camera Parameters


Notifications You must be signed in to change notification settings


Folders and files

Last commit message
Last commit date

Latest commit



10 Commits

Repository files navigation

NeRF--: Neural Radiance Fields Without Known Camera Parameters

Project Page | Latest arXiv | Colab Notebook | LLFF Released Data | BLEFF Data | ckpts

Zirui Wang¹, Shangzhe Wu², Weidi Xie², Min Chen³, Victor Adrian Prisacariu¹.

¹Active Vision Lab (AVL) + ²Visual Geometry Group (VGG) + ³e-Research Centre, University of Oxford.


8 Aug 2023

Some BLEFF-related info discussed in issues:

  • #34: training config.
  • #37: coordinate system.
  • #38: loading script.
  • #41: non-normalised rotation matrix issue.

11 Aug 2022

Release our checkpoints on LLFF dataset.

7 Apr 2022

Update our arXiv paper with

  1. A breaking point analysis for the camera parameter estimation. In short, our method can tolerate ±20 degrees of rotation variance and ±20% of translation variance.
  2. A customised dataset, which we named Blender Forward Facing (BLEFF). We will provide the dataloader file to load this data soon.

20 Apr 2021

Initial code release, corresponding to arXiv paper v3.


We provide 3 training targets in this repository, under the tasks directory:

  1. tasks/nerfmm/ This is our main training script for the NeRF-LLFF dataset, which estimates camera poses, focal lenghts and a NeRF jointly and monitors the absolute trajectory error (ATE) between our estimation of camera parameters and COLMAP estimation during training. This target can also start training from a COLMAP initialisation and refine the COLMAP camera parameters.
  2. tasks/refine_nerfmm/ This is the training script that refines a pretrained nerfmm system.
  3. tasks/any_folder/ This is a training script that takes a folder that contains forward-facing images and trains with our nerfmm system without making any comparison with COLMAP. It is similar to what we offer in our CoLab notebook and we treat this any_folder target as a playgraound, where users can try novel view synthesis by just providing an image folder and do not care how the camera parameter estimation compares with COLMAP.

For each target, we provide relevant utilities to evaluate our system. Specifically,

  • for the nerfmm target, we provide three utility files:
    • to evaluate image rendering quality on validation splits with PSNR, SSIM and LPIPS, i.e, results in Table 1.
    • to render novel views using a spiral camera trajectory, i.e. results in Figure 1.
    • to visualise our camera parameter estimation with COLMAP estimation in 3D. It also computes ATE between them, i.e. E1 in Table 2.
  • for the refine_nerfmm target, all utilities in nerfmm target above are compatible with refine_nerfmm target, since it just refines a pretrained nerfmm system.
  • for the any_folder target, it has its own and utilities, as it does not compare with COLMAP. It does not have a file as this target is treated as a playground and does not split images to train/validation sets. It only provides novel view synthesis results via the file.

Table of Content


We provide a environment.yml file to set up a conda environment:

git clone
cd nerfmm
conda env create -f environment.yml

Generally, our code should be able to run with any pytorch >= 1.1 .

(Optional) Install open3d for visualisation. You might need a physical monitor to install this lib.

pip install open3d

Get Data

LLFF Dataset

We use the LLFF dataset with two small structural changes:

  1. We remove their image_4 and image_8 folder and downsample images to any desirable resolution during data loading dataloader/, by calling PyTorch's interpolate function.
  2. We explicitly generate two txt files for train/val image ids. i.e. take every 8th image as the validation set, as in the official NeRF train/val split. The only difference is that we store them as txt files while NeRF split them during data loading. The file produces these two txt files is utils/

In addition to the NeRF-LLFF dataset, we provide two demo scenes to demonstrate how to use the any_folder target.

We pack the re-structured LLFF data and our data to a tar ball (~1.8G), to get it, run:

tar -xzvf path/to/the/tar.gz

BLEFF Dataset

There are mainly two reasons that motivate us to create BLEFF:

  1. We need to evaluate both the camera parameter estimation accuracy and image rendering quality at the same time.
  2. To facilitate the analysis of the robustness of our method, a dataset with progressively increasing pose perturbation levels is required.

To that end, we introduce a synthetic dataset BLEFF, containing 14 path-traced scenes, with each rendered in multiple levels of rotation and translation perturbations. Those scenes are modified and rendered with open-source blender files on blendswap and the license info can be found in our supplementary file.

Usage of our data:

tar -xzvf path/to/the/tar.gz


We show how to:

  1. train a nerfmm from scratch, i.e. initialise camera poses with identity matrices and focal lengths with image resolution:
    python tasks/nerfmm/ \
    --base_dir='path/to/nerfmm_release/data' \
  2. train a nerfmm from COLMAP initialisation:
    python tasks/nerfmm/ \
    --base_dir='path/to/nerfmm_release/data' \
    --scene_name='LLFF/fern' \
    --start_refine_pose_epoch=1000 \
    This command initialises a nerfmm target with COLMAP parameters, trains with them for 1000 epochs, and starts refining those parameters after 1000 epochs.
  3. train a nerfmm from a pretrained nerfmm:
    python tasks/refine_nerfmm/ \
    --base_dir='path/to/nerfmm_release/data' \
    --scene_name='LLFF/fern' --start_refine_epoch=1000 \
    This command initialises a refine_nerfmm target with a set of pretrained nerfmm parameters, trains with them for 1000 epochs, and starts refining those parameters after 1000 epochs.
  4. train an any_folder from scratch given an image folder:
    python tasks/any_folder/ \
    --base_dir='path/to/nerfmm_release/data' \
    This command trains an any_folder target using a provided demo scene desk.

(Optional) set a symlink to the downloaded data:

mkdir data_dir  # do it in this nerfmm repo
cd data_dir
ln -s /path/to/downloaded/data ./nerfmm_release_data
cd ..

this can simplify the above training commands, for example:

python tasks/nerfmm/


Compute image quality metrics

Call in nerfmm target:

python tasks/nerfmm/ \
--base_dir='path/to/nerfmm_release/data' \
--scene_name='LLFF/fern' \

This file can be used to evaluate a checkpoint trained with nerfmm or refine_nerfmm target. For some scenes, you might need to tweak with --opt_eval_lr option to get the best results. Common values for opt_eval_lr are 0.01 / 0.005 / 0.001 / 0.0005 / 0.0001. The default value is 0.001. Overall, it finds validation poses that can produce highest PSNR on validation set while freezing NeRF and focal lengths. We do this because the learned camera pose space is different from the COLMAP estimated camera pose space.

Render novel views

Call in each target. The in nerfmm is compatible with refine_nerfmm target:

python \
--base_dir='path/to/nerfmm_release/data' \
--scene_name='LLFF/fern' \

Visualise estimated poses in 3D

Call in each target. The in nerfmm is compatible with refine_nerfmm target:

python \
--base_dir='path/to/nerfmm_release/data' \
--scene_name='LLFF/fern' \


Shangzhe Wu is supported by Facebook Research. Weidi Xie is supported by Visual AI (EP/T028572/1).

The authors would like to thank Tim Yuqing Tang for insightful discussions and proofreading.

During our NeRF implementation, we referenced several open sourced NeRF implementations, and we thank their contributions. Specifically, we referenced functions from nerf and nerf-pytorch, and borrowed/modified code from nerfplusplus and nerf_pl. We especially appreciate the detailed code comments and git issue answers in nerf_pl.


  title={Ne{RF}$--$: Neural Radiance Fields Without Known Camera Parameters},
  author={Zirui Wang and Shangzhe Wu and Weidi Xie and Min Chen and Victor Adrian Prisacariu},
  journal={arXiv preprint arXiv:2102.07064},