Generating Future Observations to Estimate Grasp Success in Cluttered Environments
This work addresses the challenge of robotic grasping in cluttered settings, presenting an incremental improvement over existing methods.
The paper tackles the problem of estimating grasp success in cluttered environments by comparing a model-free approach with a model-based one that generates future observations, finding that the model-based method achieves 82% accuracy compared to 72% for the model-free approach.
End-to-end self-supervised models have been proposed for estimating the success of future candidate grasps and video predictive models for generating future observations. However, none have yet studied these two strategies side-by-side for addressing the aforementioned grasping problem. We investigate and compare a model-free approach, to estimate the success of a candidate grasp, against a model-based alternative that exploits a self-supervised learnt predictive model that generates a future observation of the gripper about to grasp an object. Our experiments demonstrate that despite the end-to-end model-free model obtaining a best accuracy of 72%, the proposed model-based pipeline yields a significantly higher accuracy of 82%.