Push to know! - Visuo-Tactile based Active Object Parameter Inference with Dual Differentiable Filtering
For robotic systems to interact with objects in dynamic environments, it is essential to perceive the physical properties of the objects such as shape, friction coefficient, mass, center of mass, and inertia. Estimating the physical properties of especially novel objects is a challenging problem, using either vision or tactile sensing. The physical object properties are not salient under static or quasi-static interactions, and often each parameter is only revealed under specific interactions, making it quite an interesting research problem. In this work, we propose a novel framework to estimate key object parameters using non-prehensile manipulation using vision and tactile sensing.
Our proposed active dual differentiable filtering (ADDF) approach as part of our framework learns the object-robot interaction during non-prehensile object push to infer the object's parameters. Our proposed method enables the robotic system to employ vision and tactile information to interactively explore a novel object via non-prehensile object push. The novel proposed N-step active formulation within the differentiable filtering facilitates efficient learning of the object-robot interaction model and during inference by selecting the next best exploratory push actions (where to push? and how to push?). We extensively evaluated our framework in simulation and real-robotic scenarios, yielding superior performance to the state-of-the-art baseline.
Figure - Problem Setup for visuo-tactile based active object parameter inference
Update in progress !!