A curated reading list of papers, datasets, and simulators for embodied vision research, covering navigation, interaction, and reasoning.
Awesome Embodied Vision is a curated GitHub repository that aggregates academic resources for embodied vision research. It provides a structured list of papers, datasets, and simulators focused on enabling AI agents to see, reason, and act in interactive 3D worlds. The project solves the problem of fragmented information by offering a single, organized entry point into this rapidly growing field.
AI researchers, PhD students, and engineers working on embodied AI, robotics, and computer vision who need a comprehensive overview of state-of-the-art literature and tools.
Developers choose this list because it is meticulously maintained by experts, follows a clear task-oriented taxonomy, and provides direct links to papers, code, and demos, saving significant literature review time compared to manual searches.
Reading list for research topics in embodied vision
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Organizes resources into specific embodied tasks like PointGoal Navigation and Vision-Language Navigation, making it easy to pinpoint relevant research without sifting through unrelated papers.
Provides chronologically ordered citations for hundreds of papers with direct links to PDFs, code, and project pages, significantly accelerating literature review for academics and practitioners.
Lists key datasets like Matterport3D and HM3D alongside simulators such as Habitat and AI2-THOR, offering a centralized hub for accessing essential tools in embodied vision research.
Encourages contributions via pull requests, as noted in the README's contribution guidelines, helping keep the list current with emerging research through crowd-sourced updates.
The list only provides links to external code and resources; users must navigate disparate repositories and websites for actual implementations, adding friction to hands-on experimentation.
As a manually curated project, it may lag behind the latest publications if community contributions slow down, requiring users to supplement with their own searches for cutting-edge work.
Focuses solely on resource aggregation without tutorials or explanatory content, making it challenging for newcomers to understand how to apply these tools in practice.