Merge branch 'master' of https://github.com/charlesq34/frustum-pointnets

kitti/kitti_object.py

@@ -3,6 +3,7 @@
 Author: Charles R. Qi
 Date: September 2017
 '''
+from __future__ import print_function
 
 import os
 import sys
@@ -14,6 +15,11 @@
 sys.path.append(os.path.join(ROOT_DIR, 'mayavi'))
 import kitti_util as utils
 
+try:
+    raw_input          # Python 2
+except NameError:
+    raw_input = input  # Python 3
+
 
 class kitti_object(object):
     '''Load and parse object data into a usable format.'''
@@ -149,13 +155,13 @@ def show_lidar_with_boxes(pc_velo, objects, calib,
     if 'mlab' not in sys.modules: import mayavi.mlab as mlab
     from viz_util import draw_lidar_simple, draw_lidar, draw_gt_boxes3d
 
-    print('All point num: ', pc_velo.shape[0])
+    print(('All point num: ', pc_velo.shape[0]))
     fig = mlab.figure(figure=None, bgcolor=(0,0,0),
         fgcolor=None, engine=None, size=(1000, 500))
     if img_fov:
         pc_velo = get_lidar_in_image_fov(pc_velo, calib, 0, 0,
             img_width, img_height)
-        print('FOV point num: ', pc_velo.shape[0])
+        print(('FOV point num: ', pc_velo.shape[0]))
     draw_lidar(pc_velo, fig=fig)
 
     for obj in objects:
@@ -203,7 +209,7 @@ def dataset_viz():
         img = dataset.get_image(data_idx)
         img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) 
         img_height, img_width, img_channel = img.shape
-        print('Image shape: ', img.shape)
+        print(('Image shape: ', img.shape))
         pc_velo = dataset.get_lidar(data_idx)[:,0:3]
         calib = dataset.get_calibration(data_idx)
 

kitti/kitti_util.py

@@ -3,6 +3,7 @@
 Author: Charles R. Qi
 Date: September 2017
 """
+from __future__ import print_function
 
 import numpy as np
 import cv2
@@ -287,7 +288,7 @@ def project_to_image(pts_3d, P):
     '''
     n = pts_3d.shape[0]
     pts_3d_extend = np.hstack((pts_3d, np.ones((n,1))))
-    print('pts_3d_extend shape: ', pts_3d_extend.shape)
+    print(('pts_3d_extend shape: ', pts_3d_extend.shape))
     pts_2d = np.dot(pts_3d_extend, np.transpose(P)) # nx3
     pts_2d[:,0] /= pts_2d[:,2]
     pts_2d[:,1] /= pts_2d[:,2]

kitti/prepare_data.py

@@ -3,6 +3,7 @@
 Author: Charles R. Qi
 Date: September 2017
 '''
+from __future__ import print_function
 
 import os
 import sys
@@ -52,7 +53,7 @@ def demo():
     img = dataset.get_image(data_idx)
     img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB) 
     img_height, img_width, img_channel = img.shape
-    print('Image shape: ', img.shape)
+    print(('Image shape: ', img.shape))
     pc_velo = dataset.get_lidar(data_idx)[:,0:3]
     calib = dataset.get_calibration(data_idx)
 
@@ -84,7 +85,7 @@ def demo():
     box3d_pts_2d, box3d_pts_3d = utils.compute_box_3d(objects[0], calib.P) 
     box3d_pts_3d_velo = calib.project_rect_to_velo(box3d_pts_3d)
     box3droi_pc_velo, _ = extract_pc_in_box3d(pc_velo, box3d_pts_3d_velo)
-    print('Number of points in 3d box: ', box3droi_pc_velo.shape[0])
+    print(('Number of points in 3d box: ', box3droi_pc_velo.shape[0]))
 
     fig = mlab.figure(figure=None, bgcolor=(0,0,0),
         fgcolor=None, engine=None, size=(1000, 500))
@@ -120,7 +121,7 @@ def demo():
         objects[0].xmin, objects[0].ymin, objects[0].xmax, objects[0].ymax
     boxfov_pc_velo = \
         get_lidar_in_image_fov(pc_velo, calib, xmin, ymin, xmax, ymax)
-    print('2d box FOV point num: ', boxfov_pc_velo.shape[0])
+    print(('2d box FOV point num: ', boxfov_pc_velo.shape[0]))
 
     fig = mlab.figure(figure=None, bgcolor=(0,0,0),
         fgcolor=None, engine=None, size=(1000, 500))
@@ -179,7 +180,7 @@ def extract_frustum_data(idx_filename, split, output_filename, viz=False,
     pos_cnt = 0
     all_cnt = 0
     for data_idx in data_idx_list:
-        print '------------- ', data_idx
+        print('------------- ', data_idx)
         calib = dataset.get_calibration(data_idx) # 3 by 4 matrix
         objects = dataset.get_label_objects(data_idx)
         pc_velo = dataset.get_lidar(data_idx)
@@ -200,8 +201,8 @@ def extract_frustum_data(idx_filename, split, output_filename, viz=False,
                 # Augment data by box2d perturbation
                 if perturb_box2d:
                     xmin,ymin,xmax,ymax = random_shift_box2d(box2d)
-                    print box2d
-                    print xmin,ymin,xmax,ymax
+                    print(box2d)
+                    print(xmin,ymin,xmax,ymax)
                 else:
                     xmin,ymin,xmax,ymax = box2d
                 box_fov_inds = (pc_image_coord[:,0]<xmax) & \
@@ -284,7 +285,7 @@ def get_box3d_dim_statistics(idx_filename):
     ry_list = []
     data_idx_list = [int(line.rstrip()) for line in open(idx_filename)]
     for data_idx in data_idx_list:
-        print '------------- ', data_idx
+        print('------------- ', data_idx)
         calib = dataset.get_calibration(data_idx) # 3 by 4 matrix
         objects = dataset.get_label_objects(data_idx)
         for obj_idx in range(len(objects)):

mayavi/viz_util.py

@@ -9,6 +9,12 @@
 import numpy as np
 import mayavi.mlab as mlab
 
+try:
+    raw_input          # Python 2
+except NameError:
+    raw_input = input  # Python 3
+
+
 def draw_lidar_simple(pc, color=None):
     ''' Draw lidar points. simplest set up. '''
     fig = mlab.figure(figure=None, bgcolor=(0,0,0), fgcolor=None, engine=None, size=(1600, 1000))

models/frustum_pointnets_v1.py

@@ -1,5 +1,6 @@
 ''' Frsutum PointNets v1 Model.
 '''
+from __future__ import print_function
 
 import sys
 import os
@@ -189,7 +190,7 @@ def get_model(point_cloud, one_hot_vec, is_training, bn_decay=None):
         inputs = tf.zeros((32,1024,4))
         outputs = get_model(inputs, tf.ones((32,3)), tf.constant(True))
         for key in outputs:
-            print(key, outputs[key])
+            print((key, outputs[key]))
         loss = get_loss(tf.zeros((32,1024),dtype=tf.int32),
             tf.zeros((32,3)), tf.zeros((32,),dtype=tf.int32),
             tf.zeros((32,)), tf.zeros((32,),dtype=tf.int32),

models/frustum_pointnets_v2.py

@@ -1,5 +1,6 @@
 ''' Frustum PointNets v2 Model.
 '''
+from __future__ import print_function
 
 import sys
 import os
@@ -173,7 +174,7 @@ def get_model(point_cloud, one_hot_vec, is_training, bn_decay=None):
         inputs = tf.zeros((32,1024,4))
         outputs = get_model(inputs, tf.ones((32,3)), tf.constant(True))
         for key in outputs:
-            print(key, outputs[key])
+            print((key, outputs[key]))
         loss = get_loss(tf.zeros((32,1024),dtype=tf.int32),
             tf.zeros((32,3)), tf.zeros((32,),dtype=tf.int32),
             tf.zeros((32,)), tf.zeros((32,),dtype=tf.int32),

models/tf_ops/3d_interpolation/tf_interpolate.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 import tensorflow as tf
 from tensorflow.python.framework import ops
 import sys
@@ -51,8 +52,8 @@ def _three_interpolate_grad(op, grad_out):
         now = time.time() 
         for _ in range(100):
             ret = sess.run(interpolated_points)
-        print time.time() - now
-        print ret.shape, ret.dtype
+        print(time.time() - now)
+        print(ret.shape, ret.dtype)
         #print ret
 
 

models/tf_ops/3d_interpolation/tf_interpolate_op_test.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 import tensorflow as tf
 import numpy as np
 from tf_interpolate import three_nn, three_interpolate
@@ -9,15 +10,15 @@ def test(self):
   def test_grad(self):
     with self.test_session():
       points = tf.constant(np.random.random((1,8,16)).astype('float32'))
-      print points
+      print(points)
       xyz1 = tf.constant(np.random.random((1,128,3)).astype('float32'))
       xyz2 = tf.constant(np.random.random((1,8,3)).astype('float32'))
       dist, idx = three_nn(xyz1, xyz2)
       weight = tf.ones_like(dist)/3.0
       interpolated_points = three_interpolate(points, idx, weight)
-      print interpolated_points
+      print(interpolated_points)
       err = tf.test.compute_gradient_error(points, (1,8,16), interpolated_points, (1,128,16))
-      print err
+      print(err)
       self.assertLess(err, 1e-4) 
 
 if __name__=='__main__':

models/tf_ops/3d_interpolation/visu_interpolation.py

@@ -1,4 +1,5 @@
 ''' Visualize part segmentation '''
+from __future__ import print_function
 import os
 import sys
 ROOT_DIR = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
@@ -23,7 +24,7 @@ def fun(xyz1,xyz2,pts2):
         dist = tf.maximum(dist, 1e-10)
         norm = tf.reduce_sum((1.0/dist),axis=2,keep_dims=True)
         norm = tf.tile(norm, [1,1,3])
-        print norm
+        print(norm)
         weight = (1.0/dist) / norm
         interpolated_points = three_interpolate(points, idx, weight)
     with tf.Session('') as sess:

models/tf_ops/grouping/test_knn.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 import tensorflow as tf
 import numpy as np
 
@@ -7,34 +8,34 @@
 a_val = np.random.random((2,5,3))
 b_val = np.random.random((2,2,3))
 for b in range(2):
-    print '--- ', b
+    print('--- ', b)
     t1 = a_val[b,:,:]
     t2 = b_val[b,:,:]
     for i in range(2): #npoint in b
-        print '-- point b: ', i
+        print('-- point b: ', i)
         for j in range(5): # npoint in a
             d = np.sum((t2[i,:]-t1[j,:])**2)
-            print d
+            print(d)
 
 
 
 a = tf.constant(a_val)
 b = tf.constant(b_val)
-print a.get_shape()
+print(a.get_shape())
 k = 3
 
 a = tf.tile(tf.reshape(a, (2,1,5,3)), [1,2,1,1])
 b = tf.tile(tf.reshape(b, (2,2,1,3)), [1,1,5,1])
 
 dist = -tf.reduce_sum((a-b)**2, -1)
-print dist
+print(dist)
 
 val, idx = tf.nn.top_k(dist, k=k)
-print val, idx
+print(val, idx)
 sess = tf.Session()
-print sess.run(a)
-print sess.run(b)
-print sess.run(dist)
-print sess.run(val)
-print sess.run(idx)
-print sess.run(idx).shape
+print(sess.run(a))
+print(sess.run(b))
+print(sess.run(dist))
+print(sess.run(val))
+print(sess.run(idx))
+print(sess.run(idx).shape)

models/tf_ops/grouping/tf_grouping.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 import tensorflow as tf
 from tensorflow.python.framework import ops
 import sys
@@ -98,8 +99,8 @@ def knn_point(k, xyz1, xyz2):
         now = time.time() 
         for _ in range(100):
             ret = sess.run(grouped_points)
-        print time.time() - now
-        print ret.shape, ret.dtype
-        print ret
+        print(time.time() - now)
+        print(ret.shape, ret.dtype)
+        print(ret)
 
 

models/tf_ops/grouping/tf_grouping_op_test.py

@@ -1,3 +1,4 @@
+from __future__ import print_function
 import tensorflow as tf
 import numpy as np
 from tf_grouping import query_ball_point, group_point
@@ -9,19 +10,19 @@ def test(self):
   def test_grad(self):
     with tf.device('/gpu:0'):
       points = tf.constant(np.random.random((1,128,16)).astype('float32'))
-      print points
+      print(points)
       xyz1 = tf.constant(np.random.random((1,128,3)).astype('float32'))
       xyz2 = tf.constant(np.random.random((1,8,3)).astype('float32'))
       radius = 0.3 
       nsample = 32
       idx, pts_cnt = query_ball_point(radius, nsample, xyz1, xyz2)
       grouped_points = group_point(points, idx)
-      print grouped_points
+      print(grouped_points)
 
     with self.test_session():
-      print "---- Going to compute gradient error"
+      print("---- Going to compute gradient error")
       err = tf.test.compute_gradient_error(points, (1,128,16), grouped_points, (1,8,32,16))
-      print err
+      print(err)
       self.assertLess(err, 1e-4) 
 
 if __name__=='__main__':

models/tf_ops/sampling/tf_sampling.py

@@ -3,6 +3,7 @@
 Modified by Charles R. Qi
 All Rights Reserved. 2017. 
 '''
+from __future__ import print_function
 import tensorflow as tf
 from tensorflow.python.framework import ops
 import sys
@@ -79,11 +80,11 @@ def farthest_point_sample(npoint,inp):
         us=(uplusv+uminusv)*0.5
         vs=(uplusv-uminusv)*0.5
         pt_sample=tria_sample+(trib_sample-tria_sample)*tf.expand_dims(us,-1)+(tric_sample-tria_sample)*tf.expand_dims(vs,-1)
-        print 'pt_sample: ', pt_sample
+        print('pt_sample: ', pt_sample)
         reduced_sample=gather_point(pt_sample,farthest_point_sample(1024,pt_sample))
-        print reduced_sample
+        print(reduced_sample)
     with tf.Session('') as sess:
         ret=sess.run(reduced_sample)
-    print ret.shape,ret.dtype
+    print(ret.shape,ret.dtype)
     import cPickle as pickle
     pickle.dump(ret,open('1.pkl','wb'),-1)

train/box_util.py

@@ -3,6 +3,7 @@
 Collected by Charles R. Qi
 Date: September 2017
 """
+from __future__ import print_function
 
 import numpy as np
 from scipy.spatial import ConvexHull
@@ -198,7 +199,7 @@ def plot_polys(plist,scale=500.0):
     points = np.random.rand(30, 2)   # 30 random points in 2-D
     hull = ConvexHull(points)
     # **In 2D "volume" is is area, "area" is perimeter
-    print('Hull area: ', hull.volume)
+    print(('Hull area: ', hull.volume))
     for simplex in hull.simplices:
         print(simplex)
 
@@ -213,7 +214,7 @@ def plot_polys(plist,scale=500.0):
     rect2 = [(150,150),(300,300),(150,450),(0,300)] 
     plot_polys([rect1, rect2])
     inter, area = convex_hull_intersection(rect1, rect2)
-    print(inter, area)
+    print((inter, area))
     if inter is not None:
         print(poly_area(np.array(inter)[:,0], np.array(inter)[:,1]))
 
@@ -230,4 +231,4 @@ def plot_polys(plist,scale=500.0):
     rect2 = [rect2[0], rect2[3], rect2[2], rect2[1]]
     plot_polys([rect1, rect2])
     inter, area = convex_hull_intersection(rect1, rect2)
-    print(inter, area)
+    print((inter, area))