Scentience provides a series of models for virtually any olfactory task. Whether you want to train a UAV to navigate by scent, interface olfaction with a robotic world model, enhance scene understanding in your VR app, or simply give olfactory awareness and chemical grounding to your LLM, the Scentience models are fully open sourced to enable rapid application development.
Scentience olfactory environments are designed for use in simulating robotics tasks or integrating into robotic world models.
These environments are designed to be used either as standalone sandboxes or integrated into popular training APIs such as MuJoCo, Gymnasium, and Nvidia IsaacSim.
Please note that all Scentience machine learning models and environments are open sourced under the MIT license and intended for research purposes only. Scentience does not claim any specific performance beyond the model and env cards nor for any specific applications.
For more information on Scentience privacy and data policies, please observe the Scentience Privacy Policy.
