Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Carla FollowLane Q-Learn with red segmentation image #172

Open
wants to merge 67 commits into
base: main
Choose a base branch
from
Open

Carla FollowLane Q-Learn with red segmentation image #172

wants to merge 67 commits into from

Conversation

pjfernandecabo
Copy link
Collaborator

@pjfernandecabo pjfernandecabo mentioned this pull request Mar 6, 2023
Closed
@pjfernandecabo pjfernandecabo changed the title fixed windows size, print messsages Carla FollowLane Q-Learn with red segmentation image Mar 6, 2023
pjfernandecabo and others added 23 commits March 9, 2023 16:50
@pjfernandecabo
first correct training
aa3b478
@pjfernandecabo
seizing PR: adding image png for README
9c6ff82
@pjfernandecabo
eliminating displaymanager
744e391
@pjfernandecabo
fixed carla server disconnection, changing displaymanager for cv2
3416e82
@pjfernandecabo
adding new sensors lane changing and obstacles, and changing to town 07
477ae79
@sergiopaniego
Updated color mask for Town07
26a2eb3
@sergiopaniego
BEV central line mask for simplified perception
034694f
@sergiopaniego
Updated BEV mask with distance to center
a989808
@pjfernandecabo
Merge pull request #180 from JdeRobot/issue-179
ec42329 
Updated color mask for semantic segmentation and BEV central line mask
@pjfernandecabo
town04 fixing rgb mask
575b446
@pjfernandecabo
adding points and info in mask image, and modifying mask threshold to…
1fb9429 
… harness red color in mask
@pjfernandecabo
training many epochs in town07, changing qlearn algorithm, distance t…
fba0077 
…o center image
@pjfernandecabo
fixing state_0, throttle and steering
975cce0
@sergiopaniego
Mask updated for Town04
29f18fc
@sergiopaniego
Merge pull request #182 from JdeRobot/issue-181
ae2ceb2 
Mask updated for Town04
@pjfernandecabo
adding alternate pose in mountain circuit Town07 with 12 poses in lan…
b5a6896 
…e -1 and 10 poses in lane 1
@pjfernandecabo
Merge branch 'carla_followlane_straight_line' of https://github.com/J…
052221a 
…deRobot/RL-Studio into carla_followlane_straight_line
@pjfernandecabo
Merge branch 'main' into carla_followlane_straight_line
d4e52f8
@pjfernandecabo
adding town04 config
20b0f9f
@pjfernandecabo
changing weakref in callbacks
f7ec80a
@pjfernandecabo
adding Carla server listener and town07_opt
f93422c
@pjfernandecabo
follow right lane only, changing rewards, fix dist variable
10a4a4e
@pjfernandecabo
changing qlearn algorithm, adding sensor collision to replicate sgeme…
9b84c23 
…ntation fault in town07
@pjfernandecabo
adding stable-baselines3: multiple algorithms and multiprocessing. Fi…
4f06f3b 
…nish action space and observation space following gym space interface
@pjfernandecabo
fixed CarlaEnv in Gym
14857ac
@pjfernandecabo
fixed sync mode YEAHH
d7cdd0d
@pjfernandecabo
sync nd reset working
5bb7af4
@pjfernandecabo
ok sync ready up to STEP()
41b1a39
@pjfernandecabo
solved sync with step
6cdb09a
@pjfernandecabo
solving DQN with perception
72ce2fd
@pjfernandecabo
solved central vs lateral states and drawing them
df64202
@pjfernandecabo
DQN sync Lane
1611bb6 
detector dev and Reset(). In Step()
@pjfernandecabo
refactoring DQN SYNC + laneDetector with regression in step() method
53c41a9
@pjfernandecabo
finish LaneDetector in step, reward and drawing
07eb0fc
@pjfernandecabo
adding step_only_right_line() method
dba7b58
@pjfernandecabo
changing LaneDetector, reset() to include new states, step() centers …
96795b3 
…with 0's
@pjfernandecabo
adding heading, function reward (center, vel, heading) and states nor…
46464bd 
…malized for DQN input
@pjfernandecabo
new recurrent DQN net with variable inputs and masking
9525bf7
@pjfernandecabo
fully integrated DQN in RL-S
0bccee4
@pjfernandecabo
fixed DQN NN with 15 states, LAneDetector perception, rewards adjusted
fc53078
@pjfernandecabo
fixed loggers. DQN First training sessions
d1deaa9
@pjfernandecabo
refactor agents/auto_carla...faster
127798c
@pjfernandecabo
refactor CarlaEnv...If garbage collector does not work...python stincks
6375956
@pjfernandecabo
adding concatenate neural net imput
f1e61de
@pjfernandecabo
killing Carla server at last
02122e0
@pjfernandecabo
adding all to run in Landau and mypc in parallel
ebfa5ab
@pjfernandecabo
adding Pythorch LaneDetector model
e396d0d
@pjfernandecabo
small changes in Landau for multi-client/multi-server
9dc9a59
@pjfernandecabo
adding collision, lane changind, obstacle sensors
ef1a4da
@pjfernandecabo
adding killing python script, 2 positions town04
e5f6220
@pjfernandecabo
dqn in both ways
08d3cdd
@pjfernandecabo
changing finish lines town04 lateral
4e6220b
@pjfernandecabo
remove an input in step
3ee0d5d
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees

@pjfernandecabo pjfernandecabo

@rubenlucas93 rubenlucas93

@sergiopaniego sergiopaniego

Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
3 participants
@pjfernandecabo @rubenlucas93 @sergiopaniego