Model ZooΒΆ

This section provides a curated collection of pre-trained models, reference implementations, and algorithm showcases developed within or integrated into the RoboOrchardLab framework. Our aim is to equip researchers and developers with ready-to-use tools to accelerate their work in embodied AI, facilitate benchmarking, and demonstrate the capabilities of our framework.

Here you will find models covering a range of tasks, including (but not limited to):

  • 3D Object Detection

  • Vision-Language Actions (VLA)

  • (Comming soon) Robotic Grasping

  • Other perception and control tasks relevant to embodied agents.

We particularly highlight algorithms and models published by our team, providing a direct pathway to reproduce and extend our research contributions.

Each model entry typically includes:

  • A brief description of the model architecture, its target task, and key features.

  • Performance metrics achieved on standard benchmarks.

  • Links to the original paper and official project page (if applicable).

  • Citation information for proper attribution.

We are continuously working to expand this collection. Please browse the available models below. Contributions from the community are also highly encouraged!