Quadruped
ForEnt
Quadruped entrapment detection and terrain traversability
Eleven forest sequences recorded with a Unitree Go2 Edu provide RGB-D, LiDAR, proprioception and third-person video with 69 annotated mobility failures.
Access links
Primary places to inspect, download, read about, or reproduce the dataset.
Quick facts
- Vehicle type
- Quadruped
- Environment
- Outdoor
- Data origin
- Real-world
- Sensor count
- 8
- Ground truth
- Not reported
- Calibration
- Parameters Only
- Annotations
- Available
Sensor overview
High-level sensor availability before the detailed sensor records below.
Ground truth
Reference scope, method, rate, coverage, and provenance.
Not reported
Calibration and synchronization
Reported calibration level, reproducibility signals, and supporting notes.
- Level
- Parameters Only
- Processed parameters
- Available
- Raw calibration data
- Not reported
- Calibration targets
- Not reported
Parameters only
Processed calibration parameters are reported, but raw calibration data is not reported.
Notes
D435i intrinsics use Kalibr with 0.66 px mean projection error. IMU noise uses Allan variance from a four-hour static recording. CAD-derived LiDAR-camera-base extrinsics are refined with Kalibr and validated by point-cloud/image alignment. Release of the static raw recording is not established.
Known or intentional limitations
Reported constraints and characteristics to check before using the dataset.
Single quadruped platform with fixed locomotion protocol; aborted sequences are retained.
Event labels begin with velocity-error candidate selection and subsequent human review.
Cameras
front_rgbd
- Model
- Intel RealSense D435i
- Setup
- Rgbd
- Count
- 1
- Effective cameras
- 2
- Modality
- Rgbd
- Resolution
- Not reported
- Rate
- 15 Hz
- Shutter
- Not Reported
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- RGB: 640x480, 69 degree HFOV, 8 ms exposure and dynamic gain. Depth: stereo IR, 85.2 degree HFOV and 58 degree VFOV; table lists range 0.3–3.0 m. RGB and depth both recorded at 15 Hz. Combined-system resolution and FOV are left null because these figures describe different streams.
third_person
- Model
- Not reported
- Setup
- Not Reported
- Count
- 1
- Effective cameras
- 1
- Modality
- Not Reported
- Resolution
- Not reported
- Rate
- Not reported
- Shutter
- Not Reported
- Lens type
- Not Reported
- HFOV
- Not reported
- VFOV
- Not reported
- Notes
- Time-aligned third-person MP4 recordings; post-processing uses sitting-to-standing motions at sequence start and end. Camera model, frame rate and physical count are not specified.
IMUs
Livox Mid-360 integrated IMU
- Model
- ICM-40609
- Count
- 1
- Accel / gyro
- Not reported
Intel RealSense D435i integrated IMU
- Model
- BMI055
- Count
- 1
- Accel / gyro
- Not reported
Unitree Go2 body IMU
- Model
- Not reported
- Count
- 1
- Accel / gyro
- Not reported
GNSS
Not reported
LiDAR
Livox Mid-360
- Model or name
- Livox Mid-360
- Count
- 1
- Dimensions
- 3
- Rate
- 10 Hz
- Channels
- Not reported
- Range
- Not reported
Radar
Not reported
Pitot tube
Not reported
Additional sensors
Odometry
- Name
- Unitree SDK kinematic odometry
- Count
- 1
- Notes
- Recorded state-estimation comparison baseline; CSV state estimates are supplied. It is not established as externally measured ground truth.
Other sensor
- Name
- Joint encoders and torques
- Count
- 1
- Notes
- Twelve actuators recorded in /lowstate at 300 Hz; this counts measurements rather than additional IMUs.
Other sensor
- Name
- Foot contact forces and battery telemetry
- Count
- 1
- Notes
- Foot forces and battery voltage/current are included in /lowstate at 300 Hz.
Other sensor
- Name
- Joystick inputs
- Count
- 1
- Notes
- Raw joystick inputs permit estimation of commanded velocities; 1 m/s is the forward command limit.
Annotations
Reported annotation availability and task support.
- Availability
- Available
- Format
- csv
- Class count
- Not reported
Semantic segmentation
Not reported
Instance segmentation
Not reported
Object detection
Not reported
Optical flow
Not reported
Depth ground truth
Not reported
Citation
Citation key and BibTeX kept at the end of the page for reference.
@misc{kirdwichai2026forent,
title={ForEnt: A Multi-Modal Dataset for Characterizing Quadruped Robot Entrapments in Forest Environments},
author={Natapat Kirdwichai and Danesh Tarapore},
year={2026},
eprint={2606.19675},
archivePrefix={arXiv},
primaryClass={cs.RO}
}
Dataset corrections
Notice a specific issue after reviewing the page?
Report an inaccuracy