GithubHelp home page GithubHelp logo

Comments (18)

javieryu avatar javieryu commented on June 27, 2024 2

Okay, so this sounds like a problem with the time synchronization in the nerfbridge ros node. This node subscribes to the pose and image topics and then converts them to the coordinate frame convention of Nerfstudio and inserts them in a pytorch dataset. The problem is that this node expects that the images and poses come in pairs with the same time stamp.

It sounds like in your case you have images and poses that do not have matching time stamps so you'll need to make changes to nsros/ros_dataloader.py. You could use an ApproximateTimeSynchronizer from message_filters see http://wiki.ros.org/message_filters#ApproximateTime_Policy.

from nerf_bridge.

cchester25 avatar cchester25 commented on June 27, 2024 1

"I solved the issue of 'ros_nerfacto' not being recognized by modifying the command line: 'python ros_train.py --method_name ros_nerfacto --data /home/pf/CODE/NERF/nerf_bridge/nsros_config_sample.json'. You should add '--method_name' before 'ros_nerfacto'."

from nerf_bridge.

cchester25 avatar cchester25 commented on June 27, 2024

"When running the command 'python ros_train.py --data /home/pf/CODE/NERF/nerf_bridge/nsros_config_sample.json' without adding 'ros_nerfacto', the code can be started, but it cannot subscribe to the image topic I published. The published topic exists and has data, and the topic name in the JSON file is consistent with the published one. What could be the reason for this?"

from nerf_bridge.

javieryu avatar javieryu commented on June 27, 2024

I'll have a look at the first part to make sure that everything is setup correctly to call the appropriate method names. As for the topic not subscribing, does the nerfbridge node time out, or is there some other behavior happening? Can you give a little more insight into the system that you are using? If your images and poses are not time synchronized then you may have to make some modifications to the nerfbridge code. I'm planning to add this as a feature when I convert the repository to ROS2.

from nerf_bridge.

cchester25 avatar cchester25 commented on June 27, 2024

I'll have a look at the first part to make sure that everything is setup correctly to call the appropriate method names. As for the topic not subscribing, does the nerfbridge node time out, or is there some other behavior happening? Can you give a little more insight into the system that you are using? If your images and poses are not time synchronized then you may have to make some modifications to the nerfbridge code. I'm planning to add this as a feature when I convert the repository to ROS2.

"I am using Ubuntu 20.04 with CPU-17-13 generation and GPU-3090ti, CUDA11.8, ROS-Noetic. Additionally, our image and pose frequencies are the same, with image topic at 15hz and pose topic also at 15hz. Below is the output during runtime, which shows that it is waiting for images."
图片

from nerf_bridge.

cchester25 avatar cchester25 commented on June 27, 2024

Okay, so this sounds like a problem with the time synchronization in the nerfbridge ros node. This node subscribes to the pose and image topics and then converts them to the coordinate frame convention of Nerfstudio and inserts them in a pytorch dataset. The problem is that this node expects that the images and poses come in pairs with the same time stamp.

It sounds like in your case you have images and poses that do not have matching time stamps so you'll need to make changes to nsros/ros_dataloader.py. You could use an ApproximateTimeSynchronizer from message_filters see http://wiki.ros.org/message_filters#ApproximateTime_Policy.
My image and posestamped timestamps are now consistent, but nerf-bridge still can't work properly, and it's stuck at the waiting for data stage.
图片

from nerf_bridge.

eiyike123 avatar eiyike123 commented on June 27, 2024

I hav

Okay, so this sounds like a problem with the time synchronization in the nerfbridge ros node. This node subscribes to the pose and image topics and then converts them to the coordinate frame convention of Nerfstudio and inserts them in a pytorch dataset. The problem is that this node expects that the images and poses come in pairs with the same time stamp.

It sounds like in your case you have images and poses that do not have matching time stamps so you'll need to make changes to nsros/ros_dataloader.py. You could use an ApproximateTimeSynchronizer from message_filters see http://wiki.ros.org/message_filters#ApproximateTime_Policy.
I am having issues with the subcriber node of Orbslam2. You didnt write thee code that creates the topic.

from nerf_bridge.

javieryu avatar javieryu commented on June 27, 2024

I've updated the documentation to solve this issue, and it will be merged with main once I've finished migrating to ROS2.

from nerf_bridge.

eiyike123 avatar eiyike123 commented on June 27, 2024

I would like to know if you have finished migrating ros2. For ros1 I couldn’t get the orslam topics

from nerf_bridge.

javieryu avatar javieryu commented on June 27, 2024

@eiyike123 It's still experimental, but yes, see the ros2 branch.

from nerf_bridge.

eiyike123 avatar eiyike123 commented on June 27, 2024

Will it work so that I can switch to
It. What about the orbslam

from nerf_bridge.

javieryu avatar javieryu commented on June 27, 2024

Will it work so that I can switch to It. What about the orbslam

Since it's still experimental, I am not sure!

from nerf_bridge.

Eric-Ho-Matrix avatar Eric-Ho-Matrix commented on June 27, 2024

Well, sorry to comment under a closed question but I have two questions highly related to this topic.

  1. I also got the question that NSROS keep waiting for image streaming. Any detailed solution to this question?
  2. Indeed, my image topic and pose topic publish as 30 Hz and 15 Hz, respectively. I know this definitely cause problem. Do you have any detailed solution to solve this. I noticed you mentioned message_filter, is there any other way to fix this?

To be clear, I use realsense d455 and Vins-fusion to get image and pose, respectively.

from nerf_bridge.

eiyike123 avatar eiyike123 commented on June 27, 2024

Well, sorry to comment under a closed question but I have two questions highly related to this topic.

  1. I also got the question that NSROS keep waiting for image streaming. Any detailed solution to this question?
  2. Indeed, my image topic and pose topic publish as 30 Hz and 15 Hz, respectively. I know this definitely cause problem. Do you have any detailed solution to solve this. I noticed you mentioned message_filter, is there any other way to fix this?

To be clear, I use realsense d455 and Vins-fusion to get image and pose, respectively.

Are you using ros1 or ros2

from nerf_bridge.

Eric-Ho-Matrix avatar Eric-Ho-Matrix commented on June 27, 2024

Well, sorry to comment under a closed question but I have two questions highly related to this topic.

  1. I also got the question that NSROS keep waiting for image streaming. Any detailed solution to this question?
  2. Indeed, my image topic and pose topic publish as 30 Hz and 15 Hz, respectively. I know this definitely cause problem. Do you have any detailed solution to solve this. I noticed you mentioned message_filter, is there any other way to fix this?

To be clear, I use realsense d455 and Vins-fusion to get image and pose, respectively.

Are you using ros1 or ros2

I'm using ros1 with ubuntu noetic.

from nerf_bridge.

eiyike123 avatar eiyike123 commented on June 27, 2024

And vin fusion publishes the pose of the images received

from nerf_bridge.

javieryu avatar javieryu commented on June 27, 2024

@Eric-Ho-Matrix Please open a new issue with a description of your problem, and I'd be happy to help you figure this out.

from nerf_bridge.

Eric-Ho-Matrix avatar Eric-Ho-Matrix commented on June 27, 2024

@Eric-Ho-Matrix Please open a new issue with a description of your problem, and I'd be happy to help you figure this out.

Sure! just open a new issue. Thanks for the help.

from nerf_bridge.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo 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.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.