Object Representations for Learning and Reasoning

Call for Papers We are sourcing short, four-page papers focusing on learning, using, and defining object representations for learning and control. Specific topics of interest include:

• Unsupervised and self-supervised object representation learning.
• Object dynamics modeling,
• Object causal and relational learning,
• Interpretability or explainability using objects,
• Applications of object representations to RL and robotics,
• Supervised and/or sim-to-real methods for learning object representations,
• Relevant connections to the notion of object studied in developmental psychology and cognitive science.

Call for Environments In addition to our call for papers, we will call for submissions of environments for testing object representations. Environment submissions should include a four page description of the environment and associated links to code repositories.

We particularly encourage submissions from students from groups that are underrepresented at machine learning conferences, including: gender, gender identity, sexual orientation, race, ethnicity, nationality, disability, and institution. We are in the process of securing sponsoring to cover registration costs for student presenters from these groups

Submission Policy:

• Submissions should be a maximum of four pages, plus any number of pages for references and supplementary material. We ask authors to use the supplementary material only for minor details that do not fit in the main paper.
• Submissions should be fully anonymized for double-blind review.
• Papers should use this style file in place of the standard NeurIPS style file.
• Dual submission policy: submissions cannot have appeared at previous conferences or at NeurIPS 2020.
• Evaluation criteria: Papers will be reviewed for topicality, clarity, correctness, and novelty; however, we encourage submissions which are cutting-edge or still work-in-progress. Environments will be evaluated for relevance to real-world tasks, importance of objects for solving the environment, (computational) resource requirements, variety of included tasks, and in particular the ease of use. Papers or environment descriptions that that are longer than four pages, clearly off-topic, or not anonymized will be rejected without review.
• Accepted submissions will appear on the workshop website (non-archival).

Important Dates and Links

UPDATE (08/11/2020): The submission deadline has been postponed to provide workshop contributors with more time.