Unsupervised Object Keypoint Learning using Local Spatial Predictability
This work provides a method for learning robust object keypoints for reinforcement learning agents, particularly in visually complex Atari environments.
This paper introduces PermaKey, a new representation learning method that identifies salient object keypoints by leveraging the predictability of local image regions from their spatial neighborhoods. The method is demonstrated on Atari, where it learns keypoints corresponding to salient object parts and improves performance on downstream RL tasks compared to competing alternatives, especially in challenging environments with moving backgrounds or distractor objects.
We propose PermaKey, a novel approach to representation learning based on object keypoints. It leverages the predictability of local image regions from spatial neighborhoods to identify salient regions that correspond to object parts, which are then converted to keypoints. Unlike prior approaches, it utilizes predictability to discover object keypoints, an intrinsic property of objects. This ensures that it does not overly bias keypoints to focus on characteristics that are not unique to objects, such as movement, shape, colour etc. We demonstrate the efficacy of PermaKey on Atari where it learns keypoints corresponding to the most salient object parts and is robust to certain visual distractors. Further, on downstream RL tasks in the Atari domain we demonstrate how agents equipped with our keypoints outperform those using competing alternatives, even on challenging environments with moving backgrounds or distractor objects.