use cnn_to_mlp to allow training of racecar using (extremely simplified) ZED camera pixel data using OpenAI baselines.

add a red sphere to make training a bit easier for now.

data/sphere2red.urdf

@@ -0,0 +1,32 @@
+<?xml version="0.0" ?>
+<robot name="urdf_robot">
+  <link name="baseLink">
+  	<contact>
+      <rolling_friction value="0.03"/>
+      <spinning_friction value="0.03"/>
+    </contact>
+
+    <inertial>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+       <mass value="10.0"/>
+       <inertia ixx="1" ixy="0" ixz="0" iyy="1" iyz="0" izz="1"/>
+    </inertial>
+    <visual>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+				<mesh filename="sphere_smooth.obj" scale="0.5 0.5 0.5"/>      
+      </geometry>
+      <material name="red">
+        <color rgba="1 0.2 0.2 1"/>
+        <specular rgb="1 1 1"/>
+      </material>
+    </visual>
+    <collision>
+      <origin rpy="0 0 0" xyz="0 0 0"/>
+      <geometry>
+	 <sphere radius="0.5"/>
+      </geometry>
+    </collision>
+  </link>
+</robot>
+

data/sphere_smooth.mtl

@@ -0,0 +1,10 @@
+# Blender MTL File: 'None'
+# Material Count: 1
+
+newmtl None
+Ns 0
+Ka 0.000000 0.000000 0.000000
+Kd 1. 0.2 0.2
+Ks 0.8 0.8 0.8
+d 1
+illum 2

examples/pybullet/gym/enjoy_pybullet_zed_racecar.py

@@ -7,7 +7,7 @@ from baselines import deepq
 def main():
     
     env = RacecarZEDGymEnv(renders=True)
-    act = deepq.load("racecar_model.pkl")
+    act = deepq.load("racecar_zed_model.pkl")
     print(act)
     while True:
         obs, done = env.reset(), False

examples/pybullet/gym/envs/bullet/racecarZEDGymEnv.py

@@ -16,7 +16,7 @@ class RacecarZEDGymEnv(gym.Env):
 
   def __init__(self,
                urdfRoot="",
-               actionRepeat=50,
+               actionRepeat=100,
                isEnableSelfCollision=True,
                renders=True):
     print("init")
@@ -24,10 +24,11 @@ class RacecarZEDGymEnv(gym.Env):
     self._urdfRoot = urdfRoot
     self._actionRepeat = actionRepeat
     self._isEnableSelfCollision = isEnableSelfCollision
-    self._observation = []
     self._ballUniqueId = -1
     self._envStepCounter = 0
     self._renders = renders
+    self._width = 100
+    self._height = 10
     self._p = p
     if self._renders:
       p.connect(p.GUI)
@@ -41,7 +42,8 @@ class RacecarZEDGymEnv(gym.Env):
     
     observation_high = np.array([np.finfo(np.float32).max] * observationDim)    
     self.action_space = spaces.Discrete(6)
-    self.observation_space = spaces.Box(-observation_high, observation_high)
+    self.observation_space = spaces.Box(low=0, high=255, shape=(self._height, self._width, 4))
+
     self.viewer = None
 
   def _reset(self):
@@ -63,7 +65,7 @@ class RacecarZEDGymEnv(gym.Env):
     bally = dist * math.cos(ang)
     ballz = 1
     
-    self._ballUniqueId = p.loadURDF("sphere2.urdf",[ballx,bally,ballz])
+    self._ballUniqueId = p.loadURDF("sphere2red.urdf",[ballx,bally,ballz])
     p.setGravity(0,0,-10)
     self._racecar = racecar.Racecar(urdfRootPath=self._urdfRoot, timeStep=self._timeStep)
     self._envStepCounter = 0
@@ -80,7 +82,6 @@ class RacecarZEDGymEnv(gym.Env):
     return [seed]
 
   def getExtendedObservation(self):
-     self._observation = [] #self._racecar.getObservation()
      carpos,carorn = p.getBasePositionAndOrientation(self._racecar.racecarUniqueId)
      carmat = p.getMatrixFromQuaternion(carorn)
      ballpos,ballorn = p.getBasePositionAndOrientation(self._ballUniqueId)      
@@ -88,25 +89,18 @@ class RacecarZEDGymEnv(gym.Env):
      ballPosInCar,ballOrnInCar = p.multiplyTransforms(invCarPos,invCarOrn,ballpos,ballorn)
      dist0 = 0.3
      dist1 = 1.
-     if self._renders:
-       print("carpos")
-       print(carpos)
-       eyePos = carpos
-       print("carmat012")
-       print(carmat[0],carmat[1],carmat[2])
-       print("carmat345")
-       print(carmat[3],carmat[4],carmat[5])
-       print("carmat678")
-       print(carmat[6],carmat[7],carmat[8])
      eyePos = [carpos[0]+dist0*carmat[0],carpos[1]+dist0*carmat[3],carpos[2]+dist0*carmat[6]+0.3]
      targetPos = [carpos[0]+dist1*carmat[0],carpos[1]+dist1*carmat[3],carpos[2]+dist1*carmat[6]+0.3]      
-      
+     up = [carmat[2],carmat[5],carmat[8]]
-       up = [0,0,1]
      viewMat = p.computeViewMatrix(eyePos,targetPos,up)
      #viewMat = p.computeViewMatrixFromYawPitchRoll(carpos,1,0,0,0,2)
-       p.getCameraImage(width=320,height=240,viewMatrix=viewMat,projectionMatrix=p.getDebugVisualizerCamera()[3],renderer=p.ER_BULLET_HARDWARE_OPENGL)
+     #print("projectionMatrix:")
-
+     #print(p.getDebugVisualizerCamera()[3])
-     self._observation.extend([ballPosInCar[0],ballPosInCar[1]])
+     projMatrix = [0.7499999403953552, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, -1.0000200271606445, -1.0, 0.0, 0.0, -0.02000020071864128, 0.0]
+     img_arr = p.getCameraImage(width=self._width,height=self._height,viewMatrix=viewMat,projectionMatrix=projMatrix)
+     rgb=img_arr[2]
+     np_img_arr = np.reshape(rgb, (self._height, self._width, 4))
+     self._observation = np_img_arr
      return self._observation
      
   def _step(self, action):

examples/pybullet/gym/train_pybullet_zed_racecar.py

@@ -14,11 +14,14 @@ def callback(lcl, glb):
     is_solved = totalt > 2000 and total >= -50
     return is_solved
 
-
 def main():
   
     env = RacecarZEDGymEnv(renders=False)
-    model = deepq.models.mlp([64])
+    model = deepq.models.cnn_to_mlp(
+        convs=[(32, 8, 4), (64, 4, 2), (64, 3, 1)],
+        hiddens=[256],
+        dueling=False
+    )
     act = deepq.learn(
         env,
         q_func=model,
@@ -30,8 +33,8 @@ def main():
         print_freq=10,
         callback=callback
     )
-    print("Saving model to racecar_model.pkl")
+    print("Saving model to racecar_zed_model.pkl")
-    act.save("racecar_model.pkl")
+    act.save("racecar_zed_model.pkl")
 
 
 if __name__ == '__main__':