policy
deepsteer-vis
pronzzz · PyTorch
or hover any field below to flag it
Overview
Name
deepsteer-vis
Author
pronzzz
Framework
PyTorch
License
MIT
Skill type
other
Evidence level
untested
Task description
DeepSteer-Vis is an autonomous driving agent trained using behavioral cloning on a modified NVIDIA PilotNet architecture with hyper-regularization. It integrates with the Udacity Simulator and features a real-time visual explainability layer using Grad-CAM to visualize the model's focus on lane line
Spaces
Action space
other · 0-dim · 0Hz
Observation space
- type: other
Links
HuggingFace repo
null
Paper (arXiv)
null
Compatible robots
0+20 mentioned but not in catalog yetNo robots list deepsteer-vis as compatible yet. Know of one? Flag it above.
Compatible environments
0No environments list deepsteer-vis yet.
Datasets that reference this policy
0No datasets reference deepsteer-vis yet.