The KITTI Vision Benchmark Suite

Method

Multi-Sensor SLAM [la] [lc] [SLAMLab]
[Anonymous Submission]

Submitted on 25 Aug. 2025 15:26 by
[Anonymous Submission]

Running time:		0.1 s
Environment:		>8 cores @ >3.5 Ghz (C/C++)

Method Description:

This is a multi-sensor SLAM method which fuses
lidar, IMU, GNSS measurements (the latter two are
not available in this benchmark). It also adds
loop closure measurements to the factor-graph
based trajectory estimation.

Demo videos can be found on YouTube:
https://www.youtube.com/@SLAMLabApp

Parameters:

clouds:
distance_filter:
enabled: true
min_distance: 7.5
max_distance: .inf
intensity_filter:
enabled: false
min_intensity: 10.0
profile_filter:
enabled: false
row_interval: 5
col_interval: 1
voxel_filter:
enabled: true
voxel_size: 0.2
normal_estimation:
enabled: false
kernel_width: 3
kernel_height: 3
max_roughness: 0.05
voxel_planes:
estimation_method: BestFit
voxel_size: 1.5
min_num_points_per_plane: 6
max_roughness: 0.30

local_map:
map_grows: true
map_radius: 50.0
min_distance_to_last_added_cloud: 0.0

correspondences:
matching_method: NearestNeighbor
max_point_to_point_distance: 1.35
max_point_to_plane_distance: 0.20
max_angle_between_normals: 30
sigma_mad_limit: 3.5

optimization:
max_iterations: 6
loss_function: TrivialLoss

Latex Bibtex:

Detailed Results

From all test sequences (sequences 11-21), our benchmark computes translational and rotational errors for all possible subsequences of length (5,10,50,100,150,...,400) meters. Our evaluation ranks methods according to the average of those values, where errors are measured in percent (for translation) and in degrees per meter (for rotation). Details for different trajectory lengths and driving speeds can be found in the plots underneath. Furthermore, the first 5 test trajectories and error plots are shown below.