
A new route to tactile sensing
Our event-driven, neuromorphic vision-based tactile technology combines displacement tracking and 3D deformation reconstruction to capture dynamic contact and micro-slip at high frequency and low latency.

ImprintX is an embodied AI company based in Beijing, with contact intelligence at its core.
We start with fine manipulation, bringing high-dimensional tactile devices, human-centered multimodal data, and models together to build general manipulation capabilities across robot embodiments.
Sense real contact. Learn precise manipulation.
From tactile perception to precise manipulation

Our event-driven, neuromorphic vision-based tactile technology combines displacement tracking and 3D deformation reconstruction to capture dynamic contact and micro-slip at high frequency and low latency.

Wearable tactile fingertips and first-person vision capture contact, force, and motion together, turning human manipulation experience into multimodal data for robot learning.

We connect tactile devices, manipulation data, and model development, improving perception, learning, and control through real-robot validation and cross-embodiment manipulation research.
Bringing together tactile perception, robot learning, and complex systems engineering

Chief Scientist · Co-founder
BAAI Scholar
Vision-based tactile sensing · Dexterous manipulation

Founder & CEO
Serial entrepreneur
Machine intelligence in the physical world · AI commercialization

CTO
Former core engineer, Tencent Search
Distributed systems · Large-scale data and systems engineering

Co-CTO · Embodied Infrastructure
PhD, Institute of Automation, CAS
Embodied AI data infrastructure · Multimodal and tactile perception

Co-CTO · Robot Learning
PhD, Institute of Automation, CAS
Robot skill learning · Multi-agent reinforcement learning
A shared mission. A long-term direction.
OUR MISSION
OUR VISION
Respect facts, expertise, and first-hand knowledge. Let the people who understand a problem guide the decision.
Discuss issues through facts and reasoning. Address disagreements openly, without making them personal.
Start with the smallest workable path. Test quickly and keep improving through real-world feedback.
Look beyond short-term trends and noise. Commit to difficult problems with lasting value.
Make decisions with facts and expertise. Move forward with candor and action.