lulinzhang / sps-nerf Goto Github PK
View Code? Open in Web Editor NEWSparseSat-NeRF: Dense Depth Supervised Neural Radiance Fields for Sparse Satellite Images
License: BSD 3-Clause "New" or "Revised" License
sps-nerf's Issues
the data of bundle adjustment
how to get the sunangles.txt and the WGS84toUTM.xml from DFC2019
About spsnerf train
I would like to express my sincere appreciation for the exceptional work you have done on your GitHub repository. Your contributions to the open-source community are truly commendable.
I am reaching out to you with a few queries that I have encountered while attempting to replicate your code.
-
During the training process, I came across a line of code that I am unable to execute:
cp Sh-SpS-Train-JAX_---_2imgs.sh "$Output"/.. I am unsure about the nature of the fileSh-SpS-Train-JAX_---_2imgs.sh. Could you please provide some insight into what this script is and how it should be used? -
Upon running the training, I encountered the following error messages:
LightningDeprecationWarning: `pytorch_lightning.metrics.*` module has been renamed to `torchmetrics.*` and split off to its own package (https://github.com/PyTorchLightning/metrics) since v1.3 and will be removed in v1.5.
RequestsDependencyWarning: urllib3 (1.26.19) or chardet (5.0.0)/charset_normalizer (2.0.12) doesn't match a supported version!
I understand that the pytorch_lightning.metrics module has been deprecated and renamed to torchmetrics. However, I am unsure about the necessary steps to update my code to avoid this deprecation warning. Could you please advise on how to resolve this issue?
I have attached the full log of the error for your reference. Your prompt response will be highly appreciated as it will help me to continue with my project.
Thank you very much for your time and assistance. I look forward to your valuable insights.
/home/harry/anaconda3/envs/spsnerf/lib/python3.6/site-packages/pytorch_lightning/metrics/__init__.py:44: LightningDeprecationWarning: `pytorch_lightning.metrics.*` module has been renamed to `torchmetrics.*` and split off to its own package (https://github.com/PyTorchLightning/metrics) since v1.3 and will be removed in v1.5
"`pytorch_lightning.metrics.*` module has been renamed to `torchmetrics.*` and split off to its own package"
/home/harry/anaconda3/envs/spsnerf/lib/python3.6/site-packages/requests/__init__.py:104: RequestsDependencyWarning: urllib3 (1.26.19) or chardet (5.0.0)/charset_normalizer (2.0.12) doesn't match a supported version!
RequestsDependencyWarning)
Running SpS_outputJAX_214-DenseDepth_ZM4-FnMd0-ds1-1 - Using gpu 0
--aoi_id: JAX_214
--beta: False
--sc_lambda: 0.0
--mapping: True
--inputdds: DenseDepth_ZM4
--ds_lambda: 1.0
--ds_drop: 1.0
--GNLL: False
--usealldepth: False
--guidedsample: True
--margin: 0.0001
--stdscale: 1.0
--corrscale: 1
--model: sps-nerf
--exp_name: SpS_outputJAX_214-DenseDepth_ZM4-FnMd0-ds1-1
--lr: 0.0005
--n_samples: 64
--n_importance: 0
------------------------------
--root_dir: /home/harry/SpS-NeRF/ProjDir/datasetJAX_214/root_dir/crops_rpcs_ba_v2/JAX_214/
--img_dir: /home/harry/SpS-NeRF/ProjDir/datasetJAX_214/JAX_214/RGB-crops/JAX_214/
--ckpts_dir: /home/harry/SpS-NeRF/ProjDir/SpS_outputJAX_214-DenseDepth_ZM4-FnMd0-ds1-1/ckpts
--logs_dir: /home/harry/SpS-NeRF/ProjDir/SpS_outputJAX_214-DenseDepth_ZM4-FnMd0-ds1-1/logs
--gt_dir: /home/harry/SpS-NeRF/ProjDir/datasetJAX_214/JAX_214/Truth
--cache_dir: /home/harry/SpS-NeRF/ProjDir/SpS_outputJAX_214-DenseDepth_ZM4-FnMd0-ds1-1/cache_dir/crops_rpcs_ba_v2/JAX_214
--ckpt_path: None
--gpu_id: 0
--batch_size: 1024
--img_downscale: 1.0
--max_train_steps: 30000
--save_every_n_epochs: 4
--fc_units: 512
--fc_layers: 8
--noise_std: 0.0
--chunk: 5120
--ds_noweights: False
--first_beta_epoch: 2
--t_embbeding_tau: 4
--t_embbeding_vocab: 30
SatNeRF: layers, feat: 8 512
GPU available: True, used: True
TPU available: False, using: 0 TPU cores
Load SatelliteRGBDEPDataset with corrscale: 1
/home/harry/SpS-NeRF/ProjDir/datasetJAX_214/root_dir/crops_rpcs_ba_v2/JAX_214//scene.loc already exist, hence skipped scaling
Image JAX_214_009_RGB loaded ( 1 / 2 )
Image JAX_214_010_RGB loaded ( 2 / 2 )
center, range: tensor([ 798962.7500, -5452430.5000, 3200708.5000]) tensor(159.)
Depth JAX_214_009_RGB loaded ( 1 / 2 )
depth range: [0.28295, 0.71207], mean: 0.59907
corr range: [0.00000, 1.00000], mean: 0.93883
std range: [0.00010, 1.00010], mean: 0.06127
74.27127 percent of pixels are valid in depth map.
段错误 (核心已转储)
About the satnerf eval results
Hello and thank you for your wonderful work!
Here I have a question for you about satnerf. I ran the code published by satnerf and tested it using the pretrained model provided, and while PSNR and SSIM were able to achieve the results reported in the article, the MAE metrics, which measure the accuracy of the DSM, differed.
I raised this issue in satnerf centreborelli/satnerf#13 but the solution he provided did not solve the problem, even the MAE = 2.900m for the Sat-NeRF model on AOI214 but reported in the article is 2.125m.I would like to ask if there is such a problem in your environment?
MicMac dense generate error
When I followed the instructions you gave me for dense_generate with micmac, I encountered some problems. This problem occurs in the section # copy the images and refined rpc parameters in 1.2 of README. An error is reported after entering the instruction: Sorry, the following FATAL ERROR happy cRPC:: ReadASCII (const std:: string & aFile) ASCII file not found.
I hope you can give me some advice. I'm stuck in this step now. Thank you!
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.