Which Physical Tactile Cues Matter for Contact-Rich Imitation Learning? A Controlled Real-Robot Study
Which Physical Tactile Cues Matter for Contact-Rich Imitation Learning? A Controlled Real-Robot Study
This paper presents a controlled real-robot study using the LeRobot platform to evaluate which physical tactile cues contribute most to contact-rich imitation learning. By deriving tactile RGB images, normal/depth maps, and marker-flow fields from a single visuo-tactile stream, the authors compare individual and combined cue contributions across 720 trials. Key findings include: (1) tactile perception improves success rates from 57.8% to 78.9%; (2) marker flow is the most effective individual cue at 74.4% success; (3) cue-factorized encoding consistently outperforms uniform processing; and (4) cross-attention fusion offers a lightweight alternative comparable to pretrained encoders.
The study provides controlled real-robot evaluation showing that marker-flow fields significantly improve contact regulation in visuo-tactile imitation learning, with tactile perception raising success rates from 57.8% to 78.9% across 720 trials.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- Controlled physical settingOpen source S3
Reported numbers
Reported success rate
57.8%
Reported trials
View original evidence
ame policy backbone, we evaluate the individual and combined contributions of these tactile cues. We further compare tactile encoders and visuo-tactile fusion strategies. Evaluation across 720 real-robot trials yields four key findings: (1) tactile perception substantially improves the overall success rate from 57.8% to 78.9%, although its contribution varies across task stages; (2) marker flow is the most effective
Open source S4Reported success rate
78.9%
Reported trials
View original evidence
ame policy backbone, we evaluate the individual and combined contributions of these tactile cues. We further compare tactile encoders and visuo-tactile fusion strategies. Evaluation across 720 real-robot trials yields four key findings: (1) tactile perception substantially improves the overall success rate from 57.8% to 78.9%, although its contribution varies across task stages; (2) marker flow is the most effective
Open source S4Reported success rate
74.4%
Reported trials
View original evidence
individual cue, achieving a success rate of 74.4% and playing a particularly important role in regulating established contact; (3) cue-factorized encoding consistently improves performance, highlighting that different physical cues require tailored processing techniques; and (4) although the effectiveness of tactile fusion is task-dependent, cross-attention offers a lightweight solution with performance comparable to
Open source S5trials
720
Reported trials
View original evidence
ame policy backbone, we evaluate the individual and combined contributions of these tactile cues. We further compare tactile encoders and visuo-tactile fusion strategies. Evaluation across 720 real-robot trials yields four key findings: (1) tactile perception substantially improves the overall success rate from 57.8% to 78.9%, although its contribution varies across task stages; (2) marker flow is the most effective
Open source S4
- paper: not_reported
What remains unknown
Not established in the collected evidence: Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
g. This paper presents a controlled real-robot study that compares physically interpretable tactile signals using the LeRobot platform, enhanced with our visuo-tactile perception system. From the same tactile stream, we derive tactile RGB images, normal/depth maps, and marker-flow fields, representing contact appearance, geometry, and tangential deformation, respectively. Using identical task demonstrations and the s
Open source S3
ame policy backbone, we evaluate the individual and combined contributions of these tactile cues. We further compare tactile encoders and visuo-tactile fusion strategies. Evaluation across 720 real-robot trials yields four key findings: (1) tactile perception substantially improves the overall success rate from 57.8% to 78.9%, although its contribution varies across task stages; (2) marker flow is the most effective
Open source S4
individual cue, achieving a success rate of 74.4% and playing a particularly important role in regulating established contact; (3) cue-factorized encoding consistently improves performance, highlighting that different physical cues require tailored processing techniques; and (4) although the effectiveness of tactile fusion is task-dependent, cross-attention offers a lightweight solution with performance comparable to
Open source S5
Source：arXiv Robotics — research abstracts · arxiv.org