ICML 2021 Workshop: Self-Supervised Learning for Reasoning and Perception


Join our workshop via Zoom: https://us02web.zoom.us/j/81362317971?pwd=WWdOWHF1VUdrRGdtZjJSME1sdGtjZz09

For poster sessions: posters can be viewed here. If interested in a poster, you can click the corresponding Zoom link. The authors are presenting their poster in their Zoom.

Online Workshop, July 24 (Saturday), 2021

Self-supervised learning (SSL) is an unsupervised approach for representation learning without relying on human-provided labels. It creates auxiliary tasks on unlabeled input data and learns representations by solving these tasks. SSL has demonstrated great success on images (e.g., MoCo, PIRL, SimCLR) and texts (e.g., BERT) and has shown promising results in other data modalities, including graphs, time-series, audio, etc. On a wide variety of tasks, SSL without using human-provided labels achieves performance that is close to fully supervised approaches.

Existing SSL research mostly focuses on perception tasks such as image classification, speech recognition, text classification, etc. SSL for reasoning tasks (e.g., symbolic reasoning on graphs, relational reasoning in computer vision, multi-hop reasoning in NLP) is largely ignored. In this workshop, we aim to bridge this gap. We bring together SSL-interested researchers from various domains to discuss how to develop SSL methods for reasoning tasks, such as how to design pretext tasks for symbolic reasoning, how to develop contrastive learning methods for relational reasoning, how to develop SSL approaches to bridge reasoning and perception, etc. Different from previous SSL-related workshops which focus on perception tasks, our workshop focuses on promoting SSL research for reasoning. The topics include but are not limited to:

  • SSL for logic and symbolic reasoning on knowledge graphs and relational data
  • SSL for relational and graph reasoning in computer vision
  • SSL for multi-hop reasoning in natural language and text corpora
  • Theoretical foundations of SSL for reasoning
  • Design of auxiliary tasks in SSL for reasoning
  • Contrastive learning for reasoning
  • SSL for bridging reasoning and perception
  • SSL for computer vision, NLP, robotics, speech, time-series analysis, graph analytics, etc.
  • SSL for healthcare, social media, neuroscience, biology, social science, etc.