Multimodal Active Measurement for Human Mesh Recovery in Close Proximity
This addresses a critical challenge for robots in physical human-robot interactions, enabling more accurate pose estimation in close-proximity scenarios, though it is incremental as it builds on existing sensor fusion and active measurement techniques.
The paper tackles the problem of poor human pose estimation accuracy in close-proximity physical human-robot interactions due to severe truncation and occlusions, by proposing an active measurement and sensor fusion framework that dynamically optimizes camera viewpoints and sensor placements and fuses reliable touch and ranging sensor measurements, resulting in outperforming previous methods on a standard occlusion benchmark and reliably estimating poses with a real robot under practical constraints.
For physical human-robot interactions (pHRI), a robot needs to estimate the accurate body pose of a target person. However, in these pHRI scenarios, the robot cannot fully observe the target person's body with equipped cameras because the target person must be close to the robot for physical interaction. This close distance leads to severe truncation and occlusions and thus results in poor accuracy of human pose estimation. For better accuracy in this challenging environment, we propose an active measurement and sensor fusion framework of the equipped cameras with touch and ranging sensors such as 2D LiDAR. Touch and ranging sensor measurements are sparse but reliable and informative cues for localizing human body parts. In our active measurement process, camera viewpoints and sensor placements are dynamically optimized to measure body parts with higher estimation uncertainty, which is closely related to truncation or occlusion. In our sensor fusion process, assuming that the measurements of touch and ranging sensors are more reliable than the camera-based estimations, we fuse the sensor measurements to the camera-based estimated pose by aligning the estimated pose towards the measured points. Our proposed method outperformed previous methods on the standard occlusion benchmark with simulated active measurement. Furthermore, our method reliably estimated human poses using a real robot, even with practical constraints such as occlusion by blankets.