Get desktop application:
View/edit binary Protocol Buffers messages
next id = 40
Used in:
,Obstacle ID
Obstacle features
from perception
Obstacle type-specific features
Obstacle tracked features
Obstacle ground-truth labels:
Obstacle short-term predicted trajectory points
Obstacle predicted trajectories
ADC trajectory at the same frame, and ADC trajectory timestamp
Surrounding lanes
self driving car intent
Used in:
Used in:
Used in:
Used in:
dim is number of masks, i.e. 12
Used in:
Features of all possible current lanes.
Features of the most possible current lane.
Features of all nearby lanes.
Lane graph
For modeling
Used in:
,Used in:
Used in:
,relative s to focus obstacle
relative l to focus obstacle
lane point heading relative to focused obstacle heading
Used in:
Used in:
s offset of the first lane point
s offset of the last lane point
If this lane-segment is the one that's closest to ADC, then adc_s gives the s of the ADC.
next id = 24
Used in:
The index of the lane-segment that is closest to ADC:
nearby obstacles on this lane sequence
right_of_way score
lane points on this lane sequence
lane points on the left neighbor lane sequence
lane points on the right neighbor lane sequence
nearby obstacles on the left neighbor lane sequence
nearby obstacles on the right neighbor lane sequence
Used in:
Used in:
Used in:
relative to focus obstacle
relative to focus obstacle
Used in:
estimated obstacle intent
Used in:
Used in:
Used in:
, ,Used in:
Used in:
, ,Used in:
Used in:
GPS time in seconds
the length of the time for this prediction (e.g. 10s)
can have multiple trajectories per obstacle
estimated obstacle intent
Feature history latest -> earliest sequence
timestamp is included in header
make prediction for multiple obstacles
perception error code
start timestamp
end timestamp
self driving car intent
Scenario
waste
Used in:
Used in:
Used in:
Used in:
Used in:
the s value within the lane
the s value along lane sequence
Used in:
,probability of this trajectory