Jv Zheng
Papers
1
Total Citations
2
H-Index
1
About
Jv Zheng is a rising researcher at the intersection of computer vision, robotics, and physical scene understanding. Their work addresses a critical gap in modern AI: while robots can recognize objects and their poses, they struggle to infer physical properties like mass, friction, and hardness from visual data alone. Zheng’s most cited work, "PUGS: Zero-Shot Physical Understanding with Gaussian Splatting" (2025), introduces a novel framework that reconstructs 3D objects using Gaussian splatting and predicts physical attributes without task-specific training. This zero-shot capability marks a significant leap toward enabling robots to interact intelligently with unfamiliar environments. Though early in their career, Zheng’s contributions are already shaping how researchers think about grounding perception in physics, bridging the gap between 3D reconstruction and real-world manipulation. Their work holds promise for applications in autonomous robotics, augmented reality, and embodied AI, where understanding an object’s weight or slipperiness is as crucial as recognizing its shape. Zheng is a name to watch in the drive toward physically aware machines.
Research Focus
Key Achievements
Top Papers
- 1PUGS: Zero-Shot Physical Understanding with Gaussian Splatting2 citations · 2025