Researching
Perception of
Autonomous Vehicles
STANFORD INTELLIGENT SYSTEMS LABORATORY
​
RESEARCH ASSISTANT
PERCEPTION for Autonomous Vehicles
JUNE 2019 - PRESENT
SKILLS​
​
-
Machine Learning with Pytorch
-
Programming in Python, Bash (Terminal)
-
Interfacing with LiDAR Datasets from CARLA, Kitti, WAYMO
-
Presenting and communicating at poster sessions
TAKEAWAYS​
​
-
Autonomy over research project, with guidance from a mentor
-
Learning cutting-edge approaches from research publications and talks
-
Collaborating with other researchers at the intersection of our works
using LiDAR data from CARLA Town 1
Shown above is a video of the occupancy maps for a car simulated in the CARLA environment.
From the left to right of the figure, we have,
(1) visualized LiDAR points (the input): blue points represent free space and red are occupied points,
(2) local occupancy map with color denoting the probability that certain areas are occupied,
(3) local variance map which provides a measure of confidence about the occupancy,
(4) global occupancy map from accumulating data about each local map area.