Relative placement tasks are an important category of tasks in which one object needs to be placed in a desired pose relative to another object. Previous work has shown success in learning relative placement tasks from just a small number of demonstrations when using relational reasoning networks with geometric inductive biases. However, such methods cannot flexibly represent multimodal tasks, like a mug hanging on any of n racks. We propose a method that incorporates additional properties that enable learning multimodal relative placement solutions, while retaining the provably translation-invariant and relational properties of prior work. We show that our method is able to learn precise relative placement tasks with only 10-20 multimodal demonstrations with no human annotations across a diverse set of objects within a category. Supplementary information can be found on the website: https://sites.google.com/view/tax-posed/home.

Qi Cheng, Mert Inan, Rahma Mbarki, Grace Grmek, Theresa Choi, Yiming Sun, Kimele Persaud, Jenny Wang, Malihe Alikhani. Proceedings of the 24th Annual Meeting of the Special Interest Group on Discourse and Dialogue. 2023.

Humans and animals explore their environment and acquire useful skills even in the absence of clear goals, exhibiting intrinsic motivation. The study of intrinsic motivation in artificial agents is concerned with the following question: what is a good general-purpose objective for an agent? We study this question in dynamic partially-observed environments, and argue that a compact and general learning objective is to minimize the entropy of the agent's state visitation estimated using a latent state-space model. This objective induces an agent to both gather information about its environment, corresponding to reducing uncertainty, and to gain control over its environment, corresponding to reducing the unpredictability of future world states. We instantiate this approach as a deep reinforcement learning agent equipped with a deep variational Bayes filter. We find that our agent learns to discover, represent, and exercise control of dynamic objects in a variety of partially-observed environments sensed with visual observations without extrinsic reward.

BACKGROUND: Scaffold proteins play a critical role in an increasing number of biological signaling processes, including simple tethering mechanism, regulating selectivity in pathways, shaping cellular behaviors. While many databases document the signaling pathways, few databases are devoted to the scaffold proteins that medicate signal transduction. RESULTS: Here, we have developed a user-friendly database, ScaPD, to describe computationally predicted, experimentally validated scaffold proteins and associated signaling pathways. It currently contains 273 scaffold proteins and 1118 associated signaling pathways. The database allows users to search, navigate and download the scaffold protein-mediated signaling networks. CONCLUSIONS: Manually curated and predicted scaffold protein data will be a foundation for further investigation of the scaffold protein in the signal transduction. With maintained up-to-date data, ScaPD ( http://bioinfo.wilmer.jhu.edu/ScaPD ) will be a valuable resource for understanding how individual signaling pathways are regulated.