InteriorGS - A high-quality 3D Gaussian semantic dataset launched by GroupCore Technology.

What is InteriorGS?

InteriorGS is a high-quality 3D Gaussian semantic dataset launched by Quncore Technology. It contains 1000 3D Gaussian semantic scenes, covering more than 80 indoor environments, such as homes, convenience stores, wedding halls, and museums. The dataset contains over 554,000 object instances across 755 categories, each with a 3D bounding box and semantic annotations, providing occupancy maps to support navigation and spatial understanding. InteriorGS is the world's first large-scale 3D dataset suitable for the free movement of intelligent agents. It uses 3D Gaussian sputtering technology to reconstruct scenes and combines large spatial models to impart semantic information. It provides rich training materials for improving the spatial perception capabilities of robots and AI agents and is publicly available on HuggingFace and GitHub for global developers.

Main functions of InteriorGS

• Diverse scenariosIt includes 1,000 3D scenes, covering more than 80 indoor environments such as homes, convenience stores, wedding banquet halls, and museums.
• High-density object annotationEach scene contains more than 554,000 object instances across 755 categories, each labeled with a 3D bounding box and semantic information.
• Map OccupationEach scenario provides an occupancy map to help the agent understand the spatial layout and support path planning and obstacle avoidance.
• Dynamic environment adaptationThe dataset enables agents to move freely in dynamic environments, enhancing their adaptability and flexibility.
• High-quality annotationBy combining 3D Gaussian sputtering technology to reconstruct the scene and imbue it with semantic information, high-quality training data can be provided for AI models.
• Large-scale datasetsAs the world's first large-scale 3D dataset suitable for the free movement of intelligent agents, it provides rich materials for model training.

InteriorGS project address

• Github repositoryhttps://github.com/manycore-research/InteriorGS
• HuggingFace model libraryhttps://huggingface.co/datasets/spatialverse/InteriorGS

Application scenarios of InteriorGS

• Indoor NavigationThe robot can navigate autonomously in complex indoor environments such as homes, offices, and shopping malls. InteriorGS provides high-precision 3D scenes and occupancy maps, helping the robot perceive its environment in real time, plan the optimal path, and avoid obstacles.
• Spatial perceptionThe 3D scenes and occupancy maps in the dataset help train the spatial perception capabilities of AI models, enabling them to better understand the layout and structure of indoor spaces.
• Virtual environment constructionInteriorGS's 3D scenes can be used to build virtual environments in virtual reality (VR) and augmented reality (AR) applications.
• Space layout optimizationArchitects and designers can use 3D scenes and occupancy maps in InteriorGS to optimize interior space layouts.
• Smart device deploymentInteriorGS data can help smart home systems better understand the indoor environment and deploy smart devices, such as smart cameras, sensors, and smart appliances, more effectively.