ctu-vras / depth_correction Goto Github PK
View Code? Open in Web Editor NEWSelf-Supervised Depth Correction of Lidar Measurements from Map Consistency Loss
License: BSD 3-Clause "New" or "Revised" License
Self-Supervised Depth Correction of Lidar Measurements from Map Consistency Loss
License: BSD 3-Clause "New" or "Revised" License
Dear authors,
After downloading the fee_corridor data set, I notice that there are two folders of point clouds in sequences. One is named as ouster_points
and the other is static_outster_points
, the former on consists of around 100 frames while the later one consists of around 30 frames. I'm wondering what are the differences between these two folders. Do they come from different environments, or simply the way they sampled are different. And if the former one is sampled when the vehicle is moving, there might be the distortions of the points, then how could we deal with such case?
Thanks!
Hi~ RuslanAgishev and Tomas Petricek:
I have to say that it is a fanncy work with intensive workload. But when I tried to learn from it and followed the instruction of doc/install.md, I just found that the https://github.com/tpet/data.git link failed. Can you re-upload it? And what is the use of this data repo?
Looking forward for your reply!
Hi, sorry for bothering again.
From the paper I notice that the reconstructed scene is skewed upwards (shown below)
However, after I run your code, I notice that the scene is actually skewed downwards, which is similar to the figure(shown below) in this paper. And I agree that
The map is bending depending on which wall the robot was closest to while driving
As a result, I am wondering why the graph in the paper is skewed upwards, or if I did not understand correctly.
By the way, I will greatly appreciate it if you could upload the weights and configs of the hyperparameters in the near future, so that I can do some comparative experiments.
Looking forward to your reply.
Thanks
Hi, sorry for bothering. After I look closer at your work, I have a question for the pipeline designed.
The pipeline provided in paper and Google drive slides
are slightly different (pipelines are shown below).
In the paper, I notice that there's one more line (connecting scan
and SLAM
) added in the pipeline, and there seems to be no such implementation in the codes provided. Thus, I think the pipeline in the slides are more accurate. I am wondering which pipeline is designed for this work.
Dear authors,
Sorry for bothering again, I am wondering will you release some of the hyperparameters when training Kitti360 dataset? To be more specific, those are used in preprocessing of the Kitti360 dataset (shown in the figure below).
It would really great if you could provide these hyperparameters! Looking forward to your reply.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.