mit-spark / kimera-multi-data Goto Github PK

Hi, thank you for providing the dataset. I have submitted several requests for it but have not yet received it via email. Could you please check the list of requests?

Hello! Thanks for sharing wonderful dataset, Kimera-multi dataset!
I have some question about dataset.

May I ask how you set the time sync to make the Kimera multi dataset?
Did you use GNSS or NTP?
I would appreciate it if you could explain it from the perspective of inter and intra!

Hello, thank you for this dataset.

I've been trying to test this dataset on multi-robot SLAM systems outside of Kimera-Multi, and I had some questions about the dataset:

1. After performing trajectory alignment with the ground-truth, should that result have its Roll angle adjusted by 90 degrees? From another issue, one of the author does confirm it's about 90 degrees, but I noticed it depends on the robot, and sometimes I also have to subtract 270 degrees from the yaw to get accurate results. Adjusting this significantly affects the RPE performance.
2. Is it okay to fast forward times in the bag files? I noticed, especially for 1208 campus hybrid, the robots stay still for a long time at the beginning and end and I was wondering if these were necessary? For example, I'm trying to use ORB-SLAM3 as one of the front-ends in a multi-robot setup and it's recommended to have IMU excitation in the beginning frames.
3. Considering this is a very large dataset, do you think there would be communication bandwidths issues for centralized systems? I noticed experiments with individual robots can get good results, but with all 8 (on 1208 for example) some of the trajectories aren't so good, so I was thinking of running sequentially.

Thank you

I tried to use tow trajectory data from the outdoor type, the problem is I dont know the starting coordinate for each. They start from the same coordinate as shown in the figure below (they look they are transected, but actually they are not), so How to make them each start from its origin coordinate to look as their gt traject

ory??

Hi,
I send request to download the dataset but I didn't receive anything?

Thanks for your brilliant work!
I have two questions to ask. Could you help me?

First, I cannot find the extrinsic between LiDAR and IMU. Also, I don't understand the extrinsic between IMU and RGB camera, IMU from D455 was used in the dataset, does that mean the extrinsic between IMU and RGB camera can be considered as eye(4)?

Thank you very much!

Hi, I send request to get the dataset for a long time, but i did not receive the email or get any reply.

Thanks for sharing your work!
I downloaded the Campus-Outdoor data but found that the bag files contain mismatched frame_id for color/depth/infra frames. It's strange that every frame_id seems to start with "acl_jackal2". But frame_ids in different files should have started with their own file names like hathor/thoth. Is this a mistake or I missed something?
The command I used is like:

rostopic echo /acl_jackal/forward/color/image_raw/compressed -b ~/bags/10_14_acl_jackal.bag --noarr

Thanks for the useful datasets. I found that all extrinsic matrices are provided except LiDAR. Is it valid? If yes, could you please provide it? Thanks a lot!

when apply them in rivz all paths start from the same point, how should run them and keeping each robot from it's own start point?

Firstly, I have been working with the Campus-Outdoor data set for testing with Kimera-Multi. I've followed the instructions provided, running the command CATKIN_WS=<path-to-catkin-ws> DATA_PATH=<path-to-campus-outdoor-data-folder> LOG_DIR=<path-to-log-folder> tmuxp load 1014-example.yaml. Despite the process completing without any module errors, I am unable to visualize the robot's path in RViz. I have tried solutions mentioned in closed issues #3, #6, and #8 on your repository, but the problem persists. I suspect an issue with the DPGO module, possibly related to loop closure detection. I have attached the log file from this run for your reference.

Secondly, I have created a custom YAML file, 1014-example-2robots.yaml, based on the original file logic, modifying it to operate with only two robots instead of six. However, I encounter an error during the execution of the second robot ( Total inter-robot loop closures: 0; DistributedLoopClosure.cpp:100] Check failed: robot_id >= config_.my_id_ (0 vs. 1)), which seems unexpected given the logical nature of the modifications. I am attaching this YAML file as well in the zip file and would greatly appreciate your insights into whether my approach is correct and what might be causing this error.

Thank you for your time and consideration.
klog.zip