diff --git a/app_base.py b/app_base.py
index 255855c22..f90755ab0 100644
--- a/app_base.py
+++ b/app_base.py
@@ -1,4 +1,4 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 class BaseApp(object):
     '''Base App class.'''
diff --git a/caffevis/app.py b/caffevis/app.py
index 582da50f8..5a34e8e5b 100644
--- a/caffevis/app.py
+++ b/caffevis/app.py
@@ -6,17 +6,17 @@
 import cv2
 import numpy as np
 import time
-import StringIO
+import io
 
 from misc import WithTimer
 from numpy_cache import FIFOLimitedArrayCache
 from app_base import BaseApp
 from image_misc import norm01, norm01c, norm0255, tile_images_normalize, ensure_float01, tile_images_make_tiles, ensure_uint255_and_resize_to_fit, get_tiles_height_width, get_tiles_height_width_ratio
 from image_misc import FormattedString, cv2_typeset_text, to_255
-from caffe_proc_thread import CaffeProcThread
-from jpg_vis_loading_thread import JPGVisLoadingThread
-from caffevis_app_state import CaffeVisAppState
-from caffevis_helper import get_pretty_layer_name, read_label_file, load_sprite_image, load_square_sprite_image, check_force_backward_true
+from .caffe_proc_thread import CaffeProcThread
+from .jpg_vis_loading_thread import JPGVisLoadingThread
+from .caffevis_app_state import CaffeVisAppState
+from .caffevis_helper import get_pretty_layer_name, read_label_file, load_sprite_image, load_square_sprite_image, check_force_backward_true
 
 
 
@@ -25,7 +25,7 @@ class CaffeVisApp(BaseApp):
 
     def __init__(self, settings, key_bindings):
         super(CaffeVisApp, self).__init__(settings, key_bindings)
-        print 'Got settings', settings
+        print('Got settings', settings)
         self.settings = settings
         self.bindings = key_bindings
 
@@ -46,10 +46,10 @@ def __init__(self, settings, key_bindings):
         import caffe
         if settings.caffevis_mode_gpu:
             caffe.set_mode_gpu()
-            print 'CaffeVisApp mode (in main thread):     GPU'
+            print('CaffeVisApp mode (in main thread):     GPU')
         else:
             caffe.set_mode_cpu()
-            print 'CaffeVisApp mode (in main thread):     CPU'
+            print('CaffeVisApp mode (in main thread):     CPU')
         self.net = caffe.Classifier(
             settings.caffevis_deploy_prototxt,
             settings.caffevis_network_weights,
@@ -58,7 +58,7 @@ def __init__(self, settings, key_bindings):
             raw_scale = self._range_scale,
         )
 
-        if isinstance(settings.caffevis_data_mean, basestring):
+        if isinstance(settings.caffevis_data_mean, str):
             # If the mean is given as a filename, load the file
             try:
 
@@ -66,7 +66,7 @@ def __init__(self, settings, key_bindings):
                 if file_extension == ".npy":
                     # load mean from numpy array
                     self._data_mean = np.load(settings.caffevis_data_mean)
-                    print "Loaded mean from numpy file, data_mean.shape: ", self._data_mean.shape
+                    print("Loaded mean from numpy file, data_mean.shape: ", self._data_mean.shape)
 
                 elif file_extension == ".binaryproto":
 
@@ -76,27 +76,27 @@ def __init__(self, settings, key_bindings):
                     blob.ParseFromString(data)
                     self._data_mean = np.array(caffe.io.blobproto_to_array(blob))
                     self._data_mean = np.squeeze(self._data_mean)
-                    print "Loaded mean from binaryproto file, data_mean.shape: ", self._data_mean.shape
+                    print("Loaded mean from binaryproto file, data_mean.shape: ", self._data_mean.shape)
 
                 else:
                     # unknown file extension, trying to load as numpy array
                     self._data_mean = np.load(settings.caffevis_data_mean)
-                    print "Loaded mean from numpy file, data_mean.shape: ", self._data_mean.shape
+                    print("Loaded mean from numpy file, data_mean.shape: ", self._data_mean.shape)
 
             except IOError:
-                print '\n\nCound not load mean file:', settings.caffevis_data_mean
-                print 'Ensure that the values in settings.py point to a valid model weights file, network'
-                print 'definition prototxt, and mean. To fetch a default model and mean file, use:\n'
-                print '$ cd models/caffenet-yos/'
-                print '$ ./fetch.sh\n\n'
+                print('\n\nCound not load mean file:', settings.caffevis_data_mean)
+                print('Ensure that the values in settings.py point to a valid model weights file, network')
+                print('definition prototxt, and mean. To fetch a default model and mean file, use:\n')
+                print('$ cd models/caffenet-yos/')
+                print('$ ./fetch.sh\n\n')
                 raise
             input_shape = self.net.blobs[self.net.inputs[0]].data.shape[-2:]   # e.g. 227x227
             # Crop center region (e.g. 227x227) if mean is larger (e.g. 256x256)
             excess_h = self._data_mean.shape[1] - input_shape[0]
             excess_w = self._data_mean.shape[2] - input_shape[1]
             assert excess_h >= 0 and excess_w >= 0, 'mean should be at least as large as %s' % repr(input_shape)
-            self._data_mean = self._data_mean[:, (excess_h/2):(excess_h/2+input_shape[0]),
-                                              (excess_w/2):(excess_w/2+input_shape[1])]
+            self._data_mean = self._data_mean[:, (excess_h//2):(excess_h//2+input_shape[0]),
+                                              (excess_w//2):(excess_w//2+input_shape[1])]
         elif settings.caffevis_data_mean is None:
             self._data_mean = None
         else:
@@ -132,7 +132,7 @@ def _populate_net_layer_info(self):
         keyboard navigation).
         '''
         self.net_layer_info = {}
-        for key in self.net.blobs.keys():
+        for key in list(self.net.blobs.keys()):
             self.net_layer_info[key] = {}
             # Conv example: (1, 96, 55, 55)
             # FC example: (1, 1000)
@@ -171,7 +171,7 @@ def get_heartbeats(self):
         return [self.proc_thread.heartbeat, self.jpgvis_thread.heartbeat]
             
     def quit(self):
-        print 'CaffeVisApp: trying to quit'
+        print('CaffeVisApp: trying to quit')
 
         with self.state.lock:
             self.state.quit = True
@@ -185,24 +185,24 @@ def quit(self):
                 raise Exception('CaffeVisApp: Could not join proc_thread; giving up.')
             self.proc_thread = None
                 
-        print 'CaffeVisApp: quitting.'
+        print('CaffeVisApp: quitting.')
         
     def _can_skip_all(self, panes):
-        return ('caffevis_layers' not in panes.keys())
+        return ('caffevis_layers' not in list(panes.keys()))
         
     def handle_input(self, input_image, panes):
         if self.debug_level > 1:
-            print 'handle_input: frame number', self.handled_frames, 'is', 'None' if input_image is None else 'Available'
+            print('handle_input: frame number', self.handled_frames, 'is', 'None' if input_image is None else 'Available')
         self.handled_frames += 1
         if self._can_skip_all(panes):
             return
 
         with self.state.lock:
             if self.debug_level > 1:
-                print 'CaffeVisApp.handle_input: pushed frame'
+                print('CaffeVisApp.handle_input: pushed frame')
             self.state.next_frame = input_image
             if self.debug_level > 1:
-                print 'CaffeVisApp.handle_input: caffe_net_state is:', self.state.caffe_net_state
+                print('CaffeVisApp.handle_input: caffe_net_state is:', self.state.caffe_net_state)
     
     def redraw_needed(self):
         return self.state.redraw_needed()
@@ -210,7 +210,7 @@ def redraw_needed(self):
     def draw(self, panes):
         if self._can_skip_all(panes):
             if self.debug_level > 1:
-                print 'CaffeVisApp.draw: skipping'
+                print('CaffeVisApp.draw: skipping')
             return False
 
         with self.state.lock:
@@ -222,7 +222,7 @@ def draw(self, panes):
 
         if do_draw:
             if self.debug_level > 1:
-                print 'CaffeVisApp.draw: drawing'
+                print('CaffeVisApp.draw: drawing')
 
             if 'caffevis_control' in panes:
                 self._draw_control_pane(panes['caffevis_control'])
@@ -321,26 +321,26 @@ def _draw_status_pane(self, pane):
                     'thick': self.settings.caffevis_status_thick}
         loc = self.settings.caffevis_status_loc[::-1]   # Reverse to OpenCV c,r order
 
-        status = StringIO.StringIO()
+        status = io.StringIO()
         fps = self.proc_thread.approx_fps()
         with self.state.lock:
-            print >>status, 'pattern' if self.state.pattern_mode else ('back' if self.state.layers_show_back else 'fwd'),
-            print >>status, '%s:%d |' % (self.state.layer, self.state.selected_unit),
+            print('pattern' if self.state.pattern_mode else ('back' if self.state.layers_show_back else 'fwd'), end=' ', file=status)
+            print('%s:%d |' % (self.state.layer, self.state.selected_unit), end=' ', file=status)
             if not self.state.back_enabled:
-                print >>status, 'Back: off',
+                print('Back: off', end=' ', file=status)
             else:
-                print >>status, 'Back: %s' % ('deconv' if self.state.back_mode == 'deconv' else 'bprop'),
-                print >>status, '(from %s_%d, disp %s)' % (self.state.backprop_layer,
+                print('Back: %s' % ('deconv' if self.state.back_mode == 'deconv' else 'bprop'), end=' ', file=status)
+                print('(from %s_%d, disp %s)' % (self.state.backprop_layer,
                                                            self.state.backprop_unit,
-                                                           self.state.back_filt_mode),
-            print >>status, '|',
-            print >>status, 'Boost: %g/%g' % (self.state.layer_boost_indiv, self.state.layer_boost_gamma)
+                                                           self.state.back_filt_mode), end=' ', file=status)
+            print('|', end=' ', file=status)
+            print('Boost: %g/%g' % (self.state.layer_boost_indiv, self.state.layer_boost_gamma), file=status)
 
             if fps > 0:
-                print >>status, '| FPS: %.01f' % fps
+                print('| FPS: %.01f' % fps, file=status)
 
             if self.state.extra_msg:
-                print >>status, '|', self.state.extra_msg
+                print('|', self.state.extra_msg, file=status)
                 self.state.extra_msg = ''
 
         strings = [FormattedString(line, defaults) for line in status.getvalue().split('\n')]
@@ -580,7 +580,7 @@ def _draw_jpgvis_pane(self, pane):
             # Some may be missing this setting
             self.settings.caffevis_jpgvis_layers
         except:
-            print '\n\nNOTE: you need to upgrade your settings.py and settings_local.py files. See README.md.\n\n'
+            print('\n\nNOTE: you need to upgrade your settings.py and settings_local.py files. See README.md.\n\n')
             raise
             
         if self.settings.caffevis_jpgvis_remap and state_layer in self.settings.caffevis_jpgvis_remap:
diff --git a/caffevis/caffe_proc_thread.py b/caffevis/caffe_proc_thread.py
index 82b4bb7dc..a1ea68764 100644
--- a/caffevis/caffe_proc_thread.py
+++ b/caffevis/caffe_proc_thread.py
@@ -4,7 +4,7 @@
 
 from codependent_thread import CodependentThread
 from misc import WithTimer
-from caffevis_helper import net_preproc_forward
+from .caffevis_helper import net_preproc_forward
 
 
 
@@ -30,7 +30,7 @@ def __init__(self, settings, net, state, loop_sleep, pause_after_keys, heartbeat
 
 
     def run(self):
-        print 'CaffeProcThread.run called'
+        print('CaffeProcThread.run called')
         frame = None
 
         import caffe
@@ -40,10 +40,10 @@ def run(self):
         # CaffeProcThread thread; it is also set in the main thread.
         if self.mode_gpu:
             caffe.set_mode_gpu()
-            print 'CaffeVisApp mode (in CaffeProcThread): GPU'
+            print('CaffeVisApp mode (in CaffeProcThread): GPU')
         else:
             caffe.set_mode_cpu()
-            print 'CaffeVisApp mode (in CaffeProcThread): CPU'
+            print('CaffeVisApp mode (in CaffeProcThread): CPU')
         
         while not self.is_timed_out():
             with self.state.lock:
@@ -110,7 +110,7 @@ def run(self):
                         try:
                             self.net.backward_from_layer(backprop_layer, diffs, zero_higher = True)
                         except AttributeError:
-                            print 'ERROR: required bindings (backward_from_layer) not found! Try using the deconv-deep-vis-toolbox branch as described here: https://github.com/yosinski/deep-visualization-toolbox'
+                            print('ERROR: required bindings (backward_from_layer) not found! Try using the deconv-deep-vis-toolbox branch as described here: https://github.com/yosinski/deep-visualization-toolbox')
                             raise
                 else:
                     with WithTimer('CaffeProcThread:deconv', quiet = self.debug_level < 1):
@@ -118,7 +118,7 @@ def run(self):
                         try:
                             self.net.deconv_from_layer(backprop_layer, diffs, zero_higher = True)
                         except AttributeError:
-                            print 'ERROR: required bindings (deconv_from_layer) not found! Try using the deconv-deep-vis-toolbox branch as described here: https://github.com/yosinski/deep-visualization-toolbox'
+                            print('ERROR: required bindings (deconv_from_layer) not found! Try using the deconv-deep-vis-toolbox branch as described here: https://github.com/yosinski/deep-visualization-toolbox')
                             raise
 
                 with self.state.lock:
@@ -135,8 +135,8 @@ def run(self):
             else:
                 time.sleep(self.loop_sleep)
         
-        print 'CaffeProcThread.run: finished'
-        print 'CaffeProcThread.run: processed %d frames fwd, %d frames back' % (self.frames_processed_fwd, self.frames_processed_back)
+        print('CaffeProcThread.run: finished')
+        print('CaffeProcThread.run: processed %d frames fwd, %d frames back' % (self.frames_processed_fwd, self.frames_processed_back))
 
     def approx_fps(self):
         '''Get the approximate frames per second processed by this
diff --git a/caffevis/caffevis_app_state.py b/caffevis/caffevis_app_state.py
index 904096d53..8690fda8c 100644
--- a/caffevis/caffevis_app_state.py
+++ b/caffevis/caffevis_app_state.py
@@ -10,13 +10,13 @@ def __init__(self, net, settings, bindings, net_layer_info):
         self.lock = Lock()  # State is accessed in multiple threads
         self.settings = settings
         self.bindings = bindings
-        self._layers = net.blobs.keys()
+        self._layers = list(net.blobs.keys())
         self._layers = self._layers[1:]  # chop off data layer
         if hasattr(self.settings, 'caffevis_filter_layers'):
             for name in self._layers:
                 if self.settings.caffevis_filter_layers(name):
-                    print '  Layer filtered out by caffevis_filter_layers: %s' % name
-            self._layers = filter(lambda name: not self.settings.caffevis_filter_layers(name), self._layers)
+                    print('  Layer filtered out by caffevis_filter_layers: %s' % name)
+            self._layers = [name for name in self._layers if not self.settings.caffevis_filter_layers(name)]
         self.net_layer_info = net_layer_info
         self.layer_boost_indiv_choices = self.settings.caffevis_boost_indiv_choices   # 0-1, 0 is noop
         self.layer_boost_gamma_choices = self.settings.caffevis_boost_gamma_choices   # 0-inf, 1 is noop
@@ -106,7 +106,7 @@ def handle_key(self, key):
             elif tag == 'pattern_mode':
                 self.pattern_mode = not self.pattern_mode
                 if self.pattern_mode and not hasattr(self.settings, 'caffevis_unit_jpg_dir'):
-                    print 'Cannot switch to pattern mode; caffevis_unit_jpg_dir not defined in settings.py.'
+                    print('Cannot switch to pattern mode; caffevis_unit_jpg_dir not defined in settings.py.')
                     self.pattern_mode = False
             elif tag == 'show_back':
                 # If in pattern mode: switch to fwd/back. Else toggle fwd/back mode
diff --git a/caffevis/caffevis_helper.py b/caffevis/caffevis_helper.py
index b2880f5bf..c054d7f5c 100644
--- a/caffevis/caffevis_helper.py
+++ b/caffevis/caffevis_helper.py
@@ -34,7 +34,7 @@ def get_pretty_layer_name(settings, layer_name):
     if hasattr(settings, 'caffevis_layer_pretty_name_fn'):
         ret = settings.caffevis_layer_pretty_name_fn(ret)
     if ret != layer_name:
-        print '  Prettified layer name: "%s" -> "%s"' % (layer_name, ret)
+        print('  Prettified layer name: "%s" -> "%s"' % (layer_name, ret))
     return ret
 
 
@@ -53,7 +53,7 @@ def crop_to_corner(img, corner, small_padding = 1, large_padding = 2):
     assert corner in (0,1,2,3), 'specify corner 0, 1, 2, or 3'
     assert img.shape[0] == img.shape[1], 'img is not square'
     assert img.shape[0] % 2 == 0, 'even number of pixels assumption violated'
-    half_size = img.shape[0]/2
+    half_size = img.shape[0]//2
     big_ii = 0 if corner in (0,1) else 1
     big_jj = 0 if corner in (0,2) else 1
     tp = small_padding + large_padding
@@ -77,14 +77,14 @@ def load_sprite_image(img_path, rows_cols, n_sprites = None):
     img = caffe_load_image(img_path, color = True, as_uint = True)
     assert img.shape[0] % rows == 0, 'sprite image has shape %s which is not divisible by rows_cols %' % (img.shape, rows_cols)
     assert img.shape[1] % cols == 0, 'sprite image has shape %s which is not divisible by rows_cols %' % (img.shape, rows_cols)
-    sprite_height = img.shape[0] / rows
-    sprite_width  = img.shape[1] / cols
+    sprite_height = img.shape[0] // rows
+    sprite_width  = img.shape[1] // cols
     sprite_channels = img.shape[2]
 
     ret = np.zeros((n_sprites, sprite_height, sprite_width, sprite_channels), dtype = img.dtype)
-    for idx in xrange(n_sprites):
+    for idx in range(n_sprites):
         # Row-major order
-        ii = idx / cols
+        ii = idx // cols
         jj = idx % cols
         ret[idx] = img[ii*sprite_height:(ii+1)*sprite_height,
                        jj*sprite_width:(jj+1)*sprite_width, :]
@@ -114,10 +114,10 @@ def check_force_backward_true(prototxt_file):
                 break
 
     if not found:
-        print '\n\nWARNING: the specified prototxt'
-        print '"%s"' % prototxt_file
-        print 'does not contain the line "force_backward: true". This may result in backprop'
-        print 'and deconv producing all zeros at the input layer. You may want to add this line'
-        print 'to your prototxt file before continuing to force backprop to compute derivatives'
-        print 'at the data layer as well.\n\n'
+        print('\n\nWARNING: the specified prototxt')
+        print('"%s"' % prototxt_file)
+        print('does not contain the line "force_backward: true". This may result in backprop')
+        print('and deconv producing all zeros at the input layer. You may want to add this line')
+        print('to your prototxt file before continuing to force backprop to compute derivatives')
+        print('at the data layer as well.\n\n')
 
diff --git a/caffevis/jpg_vis_loading_thread.py b/caffevis/jpg_vis_loading_thread.py
index 1bc69c52a..0c3a134cd 100644
--- a/caffevis/jpg_vis_loading_thread.py
+++ b/caffevis/jpg_vis_loading_thread.py
@@ -4,7 +4,7 @@
 
 from codependent_thread import CodependentThread
 from image_misc import caffe_load_image, ensure_uint255_and_resize_to_fit
-from caffevis_helper import crop_to_corner
+from .caffevis_helper import crop_to_corner
 
 
 
@@ -23,7 +23,7 @@ def __init__(self, settings, state, cache, loop_sleep, heartbeat_required):
         self.debug_level = 0
         
     def run(self):
-        print 'JPGVisLoadingThread.run called'
+        print('JPGVisLoadingThread.run called')
         
         while not self.is_timed_out():
             with self.state.lock:
@@ -69,7 +69,7 @@ def run(self):
                 img_corner = crop_to_corner(img, 2)
                 images[0] = ensure_uint255_and_resize_to_fit(img_corner, resize_shape)
             except IOError:
-                print '\nAttempted to load file %s but failed. To supress this warning, remove layer "%s" from settings.caffevis_jpgvis_layers' % (jpg_path, state_layer)
+                print('\nAttempted to load file %s but failed. To supress this warning, remove layer "%s" from settings.caffevis_jpgvis_layers' % (jpg_path, state_layer))
                 pass
 
             # 1. e.g. max_im/conv1/conv1_0037.jpg
@@ -114,4 +114,4 @@ def run(self):
                 self.state.jpgvis_to_load_key = None
                 self.state.drawing_stale = True
 
-        print 'JPGVisLoadingThread.run: finished'
+        print('JPGVisLoadingThread.run: finished')
diff --git a/codependent_thread.py b/codependent_thread.py
index 39e9ef0eb..88fa97606 100644
--- a/codependent_thread.py
+++ b/codependent_thread.py
@@ -19,7 +19,7 @@ def is_timed_out(self):
         with self.heartbeat_lock:
             now = time.time()
             if now - self.last_beat > self.heartbeat_timeout:
-                print '%s instance %s timed out after %s seconds (%s - %s = %s)' % (self.__class__.__name__, self, self.heartbeat_timeout, now, self.last_beat, now - self.last_beat)
+                print('%s instance %s timed out after %s seconds (%s - %s = %s)' % (self.__class__.__name__, self, self.heartbeat_timeout, now, self.last_beat, now - self.last_beat))
                 return True
             else:
                 return False
diff --git a/find_maxes/caffe_misc.py b/find_maxes/caffe_misc.py
index d126650bb..fbe1bf6e3 100644
--- a/find_maxes/caffe_misc.py
+++ b/find_maxes/caffe_misc.py
@@ -8,22 +8,22 @@
 def shownet(net):
     '''Print some stats about a net and its activations'''
     
-    print '%-41s%-31s%s' % ('', 'acts', 'act diffs')
-    print '%-45s%-31s%s' % ('', 'params', 'param diffs')
-    for k, v in net.blobs.items():
+    print('%-41s%-31s%s' % ('', 'acts', 'act diffs'))
+    print('%-45s%-31s%s' % ('', 'params', 'param diffs'))
+    for k, v in list(net.blobs.items()):
         if k in net.params:
             params = net.params[k]
             for pp, blob in enumerate(params):
                 if pp == 0:
-                    print '  ', 'P: %-5s'%k,
+                    print('  ', 'P: %-5s'%k, end=' ')
                 else:
-                    print ' ' * 11,
-                print '%-32s' % repr(blob.data.shape),
-                print '%-30s' % ('(%g, %g)' % (blob.data.min(), blob.data.max())),
-                print '(%g, %g)' % (blob.diff.min(), blob.diff.max())
-        print '%-5s'%k, '%-34s' % repr(v.data.shape),
-        print '%-30s' % ('(%g, %g)' % (v.data.min(), v.data.max())),
-        print '(%g, %g)' % (v.diff.min(), v.diff.max())
+                    print(' ' * 11, end=' ')
+                print('%-32s' % repr(blob.data.shape), end=' ')
+                print('%-30s' % ('(%g, %g)' % (blob.data.min(), blob.data.max())), end=' ')
+                print('(%g, %g)' % (blob.diff.min(), blob.diff.max()))
+        print('%-5s'%k, '%-34s' % repr(v.data.shape), end=' ')
+        print('%-30s' % ('(%g, %g)' % (v.data.min(), v.data.max())), end=' ')
+        print('(%g, %g)' % (v.diff.min(), v.diff.max()))
 
 
 
@@ -112,10 +112,10 @@ def convert_region(self, from_layer, to_layer, region, verbose = False):
         for ii in range(from_idx, to_idx, -1):
             converter = self.converters[ii]
             if verbose:
-                print 'pushing', self.names[ii], 'region', ret, 'through converter'
+                print('pushing', self.names[ii], 'region', ret, 'through converter')
             ret = converter(ret)
         if verbose:
-            print 'Final region at ', self.names[to_idx], 'is', ret
+            print('Final region at ', self.names[to_idx], 'is', ret)
 
         return ret
 
diff --git a/find_maxes/crop_max_patches.py b/find_maxes/crop_max_patches.py
index 20161ace9..4398cea8d 100644
--- a/find_maxes/crop_max_patches.py
+++ b/find_maxes/crop_max_patches.py
@@ -2,7 +2,7 @@
 
 import argparse
 import ipdb as pdb
-import cPickle as pickle
+import pickle as pickle
 
 from loaders import load_imagenet_mean, load_labels, caffe
 from jby_misc import WithTimer
diff --git a/find_maxes/find_max_acts.py b/find_maxes/find_max_acts.py
index c86eb2285..36e46a10f 100644
--- a/find_maxes/find_max_acts.py
+++ b/find_maxes/find_max_acts.py
@@ -2,7 +2,7 @@
 
 import argparse
 import ipdb as pdb
-import cPickle as pickle
+import pickle as pickle
 
 from loaders import load_imagenet_mean, load_labels, caffe
 from jby_misc import WithTimer
diff --git a/find_maxes/jby_misc.py b/find_maxes/jby_misc.py
index bc95f4667..dad86de23 100644
--- a/find_maxes/jby_misc.py
+++ b/find_maxes/jby_misc.py
@@ -24,17 +24,17 @@ def __enter__(self):
     def __exit__(self, *args):
         if not self.quiet:
             titlestr = self.title + ' ' if self.title else ''
-            print 'Elapsed %sreal: %.06f, sys: %.06f' % ((titlestr,) + self.elapsed())
+            print('Elapsed %sreal: %.06f, sys: %.06f' % ((titlestr,) + self.elapsed()))
 
 def misc_main():
-    print 'Running quick demo'
+    print('Running quick demo')
 
-    print ' -> Before with statement'
+    print(' -> Before with statement')
     with WithTimer('sleepy time'):
-        print '   -> About to sleep'
+        print('   -> About to sleep')
         time.sleep(.5)
-        print '   -> Done sleeping'
-    print ' -> After with statement'
+        print('   -> Done sleeping')
+    print(' -> After with statement')
         
     
 if __name__ == '__main__':
diff --git a/find_maxes/loaders.py b/find_maxes/loaders.py
index a714da701..f911ef50b 100644
--- a/find_maxes/loaders.py
+++ b/find_maxes/loaders.py
@@ -6,7 +6,7 @@
 caffe_root = '../../'  # this file is expected to be in {caffe_root}/experiments/something
 import sys
 loadpath = caffe_root + 'python_cpu'
-print '= = = CAFFE LOADER: LOADING CPU VERSION from path: %s = = =' % loadpath
+print('= = = CAFFE LOADER: LOADING CPU VERSION from path: %s = = =' % loadpath)
 sys.path.insert(0, loadpath)     # Use CPU compiled code for backprop vis/etc
 import caffe
 caffe.set_mode_cpu()
@@ -27,9 +27,9 @@ def load_trained_net(model_prototxt = None, model_weights = None):
         model_prototxt = load_dir + 'deploy_1.prototxt'
         model_weights = load_dir + 'caffe_imagenet_train_iter_450000'
 
-    print 'LOADER: loading net:'
-    print '  ', model_prototxt
-    print '  ', model_weights
+    print('LOADER: loading net:')
+    print('  ', model_prototxt)
+    print('  ', model_weights)
     net = caffe.Classifier(model_prototxt, model_weights)
     #net.set_phase_test()
 
diff --git a/find_maxes/max_tracker.py b/find_maxes/max_tracker.py
index 8bfd427de..131b3d60c 100644
--- a/find_maxes/max_tracker.py
+++ b/find_maxes/max_tracker.py
@@ -67,7 +67,7 @@ def update(self, blob, image_idx, image_class):
 
         #insertion_idx = zeros((n_channels,))
         #pdb.set_trace()
-        for ii in xrange(n_channels):
+        for ii in range(n_channels):
             idx = np.searchsorted(self.max_vals[ii], data_unroll[ii, maxes[ii]])
             if idx == 0:
                 # Smaller than all 10
@@ -128,17 +128,17 @@ def load_file_list(filelist):
 
 def scan_images_for_maxes(net, datadir, filelist, n_top):
     image_filenames, image_labels = load_file_list(filelist)
-    print 'Scanning %d files' % len(image_filenames)
-    print '  First file', os.path.join(datadir, image_filenames[0])
+    print('Scanning %d files' % len(image_filenames))
+    print('  First file', os.path.join(datadir, image_filenames[0]))
 
     tracker = NetMaxTracker(n_top = n_top)
-    for image_idx in xrange(len(image_filenames)):
+    for image_idx in range(len(image_filenames)):
         filename = image_filenames[image_idx]
         image_class = image_labels[image_idx]
         #im = caffe.io.load_image('../../data/ilsvrc12/mini_ilsvrc_valid/sized/ILSVRC2012_val_00000610.JPEG')
         do_print = (image_idx % 100 == 0)
         if do_print:
-            print '%s   Image %d/%d' % (datetime.now().ctime(), image_idx, len(image_filenames))
+            print('%s   Image %d/%d' % (datetime.now().ctime(), image_idx, len(image_filenames)))
         with WithTimer('Load image', quiet = not do_print):
             im = caffe.io.load_image(os.path.join(datadir, filename))
         with WithTimer('Predict   ', quiet = not do_print):
@@ -146,7 +146,7 @@ def scan_images_for_maxes(net, datadir, filelist, n_top):
         with WithTimer('Update    ', quiet = not do_print):
             tracker.update(net, image_idx, image_class)
 
-    print 'done!'
+    print('done!')
     return tracker
 
 
@@ -156,10 +156,10 @@ def save_representations(net, datadir, filelist, layer, first_N = None):
     if first_N is None:
         first_N = len(image_filenames)
     assert first_N <= len(image_filenames)
-    image_indices = range(first_N)
-    print 'Scanning %d files' % len(image_indices)
+    image_indices = list(range(first_N))
+    print('Scanning %d files' % len(image_indices))
     assert len(image_indices) > 0
-    print '  First file', os.path.join(datadir, image_filenames[image_indices[0]])
+    print('  First file', os.path.join(datadir, image_filenames[image_indices[0]]))
 
     indices = None
     rep = None
@@ -168,7 +168,7 @@ def save_representations(net, datadir, filelist, layer, first_N = None):
         image_class = image_labels[image_idx]
         do_print = (image_idx % 10 == 0)
         if do_print:
-            print '%s   Image %d/%d' % (datetime.now().ctime(), image_idx, len(image_indices))
+            print('%s   Image %d/%d' % (datetime.now().ctime(), image_idx, len(image_indices)))
         with WithTimer('Load image', quiet = not do_print):
             im = caffe.io.load_image(os.path.join(datadir, filename))
         with WithTimer('Predict   ', quiet = not do_print):
@@ -181,7 +181,7 @@ def save_representations(net, datadir, filelist, layer, first_N = None):
             indices[ii] = image_idx
             rep[ii] = net.blobs[layer].data[0]
 
-    print 'done!'
+    print('done!')
     return indices,rep
 
 
@@ -207,8 +207,8 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
     rc = RegionComputer()
     
     image_filenames, image_labels = load_file_list(filelist)
-    print 'Loaded filenames and labels for %d files' % len(image_filenames)
-    print '  First file', os.path.join(datadir, image_filenames[0])
+    print('Loaded filenames and labels for %d files' % len(image_filenames))
+    print('  First file', os.path.join(datadir, image_filenames[0]))
 
     num_top_in_mt = mt.max_locs.shape[1]
     assert num_top <= num_top_in_mt, 'Requested %d top images but MaxTracker contains only %d' % (num_top, num_top_in_mt)
@@ -225,7 +225,7 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
         if do_info:
             info_filename = os.path.join(unit_dir, 'info.txt')
             info_file = open(info_filename, 'w')
-            print >>info_file, '# is_conv val image_idx image_class i(if is_conv) j(if is_conv) filename'
+            print('# is_conv val image_idx image_class i(if is_conv) j(if is_conv) filename', file=info_file)
 
         # iterate through maxes from highest (at end) to lowest
         for max_idx_0 in range(num_top):
@@ -238,7 +238,7 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
             filename = image_filenames[im_idx]
             do_print = (max_idx_0 == 0)
             if do_print:
-                print '%s   Output file/image(s) %d/%d' % (datetime.now().ctime(), cc * num_top, n_total_images)
+                print('%s   Output file/image(s) %d/%d' % (datetime.now().ctime(), cc * num_top, n_total_images))
 
             if mt.is_conv:
                 # Compute the focus area of the data layer
@@ -272,12 +272,12 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
 
 
             if do_info:
-                print >>info_file, 1 if mt.is_conv else 0, '%.6f' % mt.max_vals[channel_idx, max_idx],
+                print(1 if mt.is_conv else 0, '%.6f' % mt.max_vals[channel_idx, max_idx], end=' ', file=info_file)
                 if mt.is_conv:
-                    print >>info_file, '%d %d %d %d' % tuple(mt.max_locs[channel_idx, max_idx]),
+                    print('%d %d %d %d' % tuple(mt.max_locs[channel_idx, max_idx]), end=' ', file=info_file)
                 else:
-                    print >>info_file, '%d %d' % tuple(mt.max_locs[channel_idx, max_idx]),
-                print >>info_file, filename
+                    print('%d %d' % tuple(mt.max_locs[channel_idx, max_idx]), end=' ', file=info_file)
+                print(filename, file=info_file)
 
             if not (do_maxes or do_deconv or do_deconv_norm or do_backprop or do_backprop_norm):
                 continue
@@ -293,7 +293,7 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
             else:
                 reproduced_val = net.blobs[layer].data[0,channel_idx]
             if abs(reproduced_val - recorded_val) > .1:
-                print 'Warning: recorded value %s is suspiciously different from reproduced value %s. Is the filelist the same?' % (recorded_val, reproduced_val)
+                print('Warning: recorded value %s is suspiciously different from reproduced value %s. Is the filelist the same?' % (recorded_val, reproduced_val))
 
             if do_maxes:
                 #grab image from data layer, not from im (to ensure preprocessing / center crop details match between image and deconv/backprop)
@@ -315,7 +315,7 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
                 out_arr = np.zeros((3,size_ii,size_jj), dtype='float32')
                 out_arr[:, out_ii_start:out_ii_end, out_jj_start:out_jj_end] = net.blobs['data'].diff[0,:,data_ii_start:data_ii_end,data_jj_start:data_jj_end]
                 if out_arr.max() == 0:
-                    print 'Warning: Deconv out_arr in range', out_arr.min(), 'to', out_arr.max(), 'ensure force_backward: true in prototxt'
+                    print('Warning: Deconv out_arr in range', out_arr.min(), 'to', out_arr.max(), 'ensure force_backward: true in prototxt')
                 if do_deconv:
                     with WithTimer('Save img  ', quiet = not do_print):
                         save_caffe_image(out_arr, os.path.join(unit_dir, 'deconv_%03d.png' % max_idx_0),
@@ -334,7 +334,7 @@ def output_max_patches(max_tracker, net, layer, idx_begin, idx_end, num_top, dat
                 out_arr = np.zeros((3,size_ii,size_jj), dtype='float32')
                 out_arr[:, out_ii_start:out_ii_end, out_jj_start:out_jj_end] = net.blobs['data'].diff[0,:,data_ii_start:data_ii_end,data_jj_start:data_jj_end]
                 if out_arr.max() == 0:
-                    print 'Warning: Deconv out_arr in range', out_arr.min(), 'to', out_arr.max(), 'ensure force_backward: true in prototxt'
+                    print('Warning: Deconv out_arr in range', out_arr.min(), 'to', out_arr.max(), 'ensure force_backward: true in prototxt')
                 if do_backprop:
                     with WithTimer('Save img  ', quiet = not do_print):
                         save_caffe_image(out_arr, os.path.join(unit_dir, 'backprop_%03d.png' % max_idx_0),
diff --git a/image_misc.py b/image_misc.py
index 6c940d8cc..b5d10b27c 100644
--- a/image_misc.py
+++ b/image_misc.py
@@ -1,4 +1,4 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 import cv2
 import numpy as np
@@ -80,11 +80,11 @@ def crop_to_square(frame):
     i_size,j_size = frame.shape[0],frame.shape[1]
     if j_size > i_size:
         # landscape
-        offset = (j_size - i_size) / 2
+        offset = (j_size - i_size) // 2
         return frame[:,offset:offset+i_size,:]
     else:
         # portrait
-        offset = (i_size - j_size) / 2
+        offset = (i_size - j_size) // 2
         return frame[offset:offset+j_size,:,:]
 
 
@@ -419,7 +419,7 @@ def cv2_typeset_text(data, lines, loc, between = ' ', string_spacing = 0, line_s
                 if fs.align == 'right':
                     locx += fs.width - boxsize[0]
                 elif fs.align == 'center':
-                    locx += (fs.width - boxsize[0])/2
+                    locx += (fs.width - boxsize[0])//2
             #print 'right boundary is', locx + boxsize[0], '(%s)' % fs.string
                     #                print 'HERE'
             right_edge = locx + boxsize[0]
diff --git a/input_fetcher.py b/input_fetcher.py
index 95eb7a9fd..0370fa1f5 100644
--- a/input_fetcher.py
+++ b/input_fetcher.py
@@ -53,28 +53,28 @@ def __init__(self, settings):
         
     def bind_camera(self):
         # Due to OpenCV limitations, this should be called from the main thread
-        print 'InputImageFetcher: bind_camera starting'
+        print('InputImageFetcher: bind_camera starting')
         if self.no_cam_present:
-            print 'InputImageFetcher: skipping camera bind (device: None)'
+            print('InputImageFetcher: skipping camera bind (device: None)')
         else:
             self.bound_cap_device = cv2.VideoCapture(self.capture_device)
             if self.bound_cap_device.isOpened():
-                print 'InputImageFetcher: capture device %s is open' % self.capture_device
+                print('InputImageFetcher: capture device %s is open' % self.capture_device)
             else:
-                print '\n\nWARNING: InputImageFetcher: capture device %s failed to open! Camera will not be available!\n\n' % self.capture_device
+                print('\n\nWARNING: InputImageFetcher: capture device %s failed to open! Camera will not be available!\n\n' % self.capture_device)
                 self.bound_cap_device = None
                 self.no_cam_present = True
-        print 'InputImageFetcher: bind_camera finished'
+        print('InputImageFetcher: bind_camera finished')
 
     def free_camera(self):
         # Due to OpenCV limitations, this should be called from the main thread
         if self.no_cam_present:
-            print 'InputImageFetcher: skipping camera free (device: None)'
+            print('InputImageFetcher: skipping camera free (device: None)')
         else:
-            print 'InputImageFetcher: freeing camera'
+            print('InputImageFetcher: freeing camera')
             del self.bound_cap_device  # free the camera
             self.bound_cap_device = None
-            print 'InputImageFetcher: camera freed'
+            print('InputImageFetcher: camera freed')
 
     def set_mode_static(self):
         with self.lock:
@@ -83,7 +83,7 @@ def set_mode_static(self):
     def set_mode_cam(self):
         with self.lock:
             if self.no_cam_present:
-                print 'WARNING: ignoring set_mode_cam, no cam present'
+                print('WARNING: ignoring set_mode_cam, no cam present')
             else:
                 self.static_file_mode = False
                 assert self.bound_cap_device != None, 'Call bind_camera first'
@@ -137,7 +137,7 @@ def run(self):
             time.sleep(self.sleep_after_read_frame)
             #print 'Reading one frame took', time.time() - start_time
 
-        print 'InputImageFetcher: exiting run method'
+        print('InputImageFetcher: exiting run method')
         #print 'InputImageFetcher: read', self.read_frames, 'frames'
 
     def get_frame(self):
diff --git a/keys.py b/keys.py
index 6293604a3..28c44240d 100644
--- a/keys.py
+++ b/keys.py
@@ -68,9 +68,9 @@
     if key_code in key_patterns[mask]:
         old_label = key_patterns[mask][code]
         if old_label != label:
-            print 'Warning: key_patterns[%s][%s] old value %s being overwritten with %s' % (mask, key_code, old_label, label)
+            print('Warning: key_patterns[%s][%s] old value %s being overwritten with %s' % (mask, key_code, old_label, label))
     if key_code != (key_code & mask):
-        print 'Warning: key_code %s for key label %s will never trigger using mask %s' % (key_code, label, mask)
+        print('Warning: key_code %s for key label %s will never trigger using mask %s' % (key_code, label, mask))
     key_patterns[mask][key_code] = label
     #if not label in key_patterns[mask]:
     #    key_patterns[mask][label] = set()
diff --git a/live_vis.py b/live_vis.py
index 9dde3be99..62e138ac0 100644
--- a/live_vis.py
+++ b/live_vis.py
@@ -1,4 +1,4 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 import sys
 import importlib
@@ -11,7 +11,7 @@
 try:
     import cv2
 except ImportError:
-    print 'Error: Could not import cv2, please install it first.'
+    print('Error: Could not import cv2, please install it first.')
     raise
 
 from misc import WithTimer
@@ -53,12 +53,12 @@ def __init__(self, settings):
 
         for module_path, app_name in settings.installed_apps:
             module = importlib.import_module(module_path)
-            print 'got module', module
+            print('got module', module)
             app_class  = getattr(module, app_name)
-            print 'got app', app_class
+            print('got app', app_class)
             self.app_classes[app_name] = app_class
 
-        for app_name, app_class in self.app_classes.iteritems():
+        for app_name, app_class in self.app_classes.items():
             app = app_class(settings, self.bindings)
             self.apps[app_name] = app
         self.help_mode = False
@@ -102,7 +102,7 @@ def init_window(self):
                                      (max_i,max_j,1))
         #print 'BUFFER IS:', self.window_buffer.shape, self.window_buffer.min(), self.window_buffer.max()
 
-        for _,pane in self.panes.iteritems():
+        for _,pane in self.panes.items():
             pane.data = self.window_buffer[pane.i_begin:pane.i_end, pane.j_begin:pane.j_end]
 
         # Allocate help pane
@@ -125,8 +125,8 @@ def run_loop(self):
         self.input_updater.start()
 
         heartbeat_functions = [self.input_updater.heartbeat]
-        for app_name, app in self.apps.iteritems():
-            print 'Starting app:', app_name
+        for app_name, app in self.apps.items():
+            print('Starting app:', app_name)
             app.start()
             heartbeat_functions.extend(app.get_heartbeats())
 
@@ -178,13 +178,13 @@ def run_loop(self):
                 key,do_redraw = self.handle_key_pre_apps(key)
                 redraw_needed |= do_redraw
                 imshow_needed |= do_redraw
-                for app_name, app in self.apps.iteritems():
+                for app_name, app in self.apps.items():
                     with WithTimer('%s:handle_key' % app_name, quiet = self.debug_level < 1):
                         key = app.handle_key(key, self.panes)
                 key = self.handle_key_post_apps(key)
                 if self.quit:
                     break
-            for app_name, app in self.apps.iteritems():
+            for app_name, app in self.apps.items():
                 redraw_needed |= app.redraw_needed()
 
             # Grab latest frame from input_updater thread
@@ -204,7 +204,7 @@ def run_loop(self):
                                since_keypress >= self.settings.keypress_pause_handle_iterations)
             if frame_for_apps is not None and do_handle_input:
                 # Pass frame to apps for processing
-                for app_name, app in self.apps.iteritems():
+                for app_name, app in self.apps.items():
                     with WithTimer('%s:handle_input' % app_name, quiet = self.debug_level < 1):
                         app.handle_input(latest_frame_data, self.panes)
                 frame_for_apps = None
@@ -214,7 +214,7 @@ def run_loop(self):
                          (since_keypress >= self.settings.keypress_pause_redraw_iterations or
                           since_redraw >= self.settings.redraw_at_least_every))
             if redraw_needed and do_redraw:
-                for app_name, app in self.apps.iteritems():
+                for app_name, app in self.apps.items():
                     with WithTimer('%s:draw' % app_name, quiet = self.debug_level < 1):
                         imshow_needed |= app.draw(self.panes)
                 redraw_needed = False
@@ -224,8 +224,8 @@ def run_loop(self):
             if imshow_needed:
                 # Only redraw pane debug if display will be updated
                 if hasattr(self.settings, 'debug_window_panes') and self.settings.debug_window_panes:
-                    for pane_name,pane in self.panes.iteritems():
-                        print pane_name, pane
+                    for pane_name,pane in self.panes.items():
+                        print(pane_name, pane)
                         pane.data[:] = pane.data * .5
                         line = [FormattedString('%s |' % pane_name, self.debug_pane_defaults),
                                 FormattedString('pos: %d,%d |' % (pane.i_begin, pane.j_begin), self.debug_pane_defaults),
@@ -255,7 +255,7 @@ def run_loop(self):
             # Extra sleep just for debugging. In production all main loop sleep should be in cv2.waitKey.
             #time.sleep(2)
 
-        print '\n\nTrying to exit run_loop...'
+        print('\n\nTrying to exit run_loop...')
         self.input_updater.quit = True
         self.input_updater.join(.01 + float(self.settings.input_updater_sleep_after_read_frame) * 5)
         if self.input_updater.is_alive():
@@ -263,11 +263,11 @@ def run_loop(self):
         else:
             self.input_updater.free_camera()
 
-        for app_name, app in self.apps.iteritems():
-            print 'Quitting app:', app_name
+        for app_name, app in self.apps.items():
+            print('Quitting app:', app_name)
             app.quit()
 
-        print 'Input thread joined and apps quit; exiting run_loop.'
+        print('Input thread joined and apps quit; exiting run_loop.')
 
     def handle_key_pre_apps(self, key):
         tag = self.bindings.get_tag(key)
@@ -289,10 +289,10 @@ def handle_key_pre_apps(self, key):
             self.help_mode = not self.help_mode
         elif tag == 'stretch_mode':
             self.input_updater.toggle_stretch_mode()
-            print 'Stretch mode is now', self.input_updater.static_file_stretch_mode
+            print('Stretch mode is now', self.input_updater.static_file_stretch_mode)
         elif tag == 'debug_level':
             self.debug_level = (self.debug_level + 1) % 3
-            for app_name, app in self.apps.iteritems():
+            for app_name, app in self.apps.items():
                 app.set_debug(self.debug_level)
         else:
             return key, False
@@ -308,12 +308,12 @@ def handle_key_post_apps(self, key):
             key_label, masked_vals = self.bindings.get_key_label_from_keycode(key, extra_info = True)
             masked_vals_pp = ', '.join(['%d (%s)' % (mv, hex(mv)) for mv in masked_vals])
             if key_label is None:
-                print 'Got key code %d (%s), did not match any known key (masked vals tried: %s)' % (key, hex(key), masked_vals_pp)
+                print('Got key code %d (%s), did not match any known key (masked vals tried: %s)' % (key, hex(key), masked_vals_pp))
             elif tag is None:
-                print 'Got key code %d (%s), matched key "%s", but key is not bound to any function' % (key, hex(key), key_label)
+                print('Got key code %d (%s), matched key "%s", but key is not bound to any function' % (key, hex(key), key_label))
             else:
-                print 'Got key code %d (%s), matched key "%s", bound to "%s", but nobody handled "%s"' % (
-                    key, hex(key), key_label, tag, tag)
+                print('Got key code %d (%s), matched key "%s", bound to "%s", but nobody handled "%s"' % (
+                    key, hex(key), key_label, tag, tag))
 
     def display_frame(self, frame):
         if self.settings.static_files_input_mode == "siamese_image_list":
@@ -350,10 +350,10 @@ def draw_help(self):
         locy = cv2_typeset_text(self.help_pane.data, lines, loc,
                                 line_spacing = self.settings.help_line_spacing)
 
-        for app_name, app in self.apps.iteritems():
+        for app_name, app in self.apps.items():
             locy = app.draw_help(self.help_pane, locy)
 
 
 
 if __name__ == '__main__':
-    print 'You probably want to run ./run_toolbox.py instead.'
+    print('You probably want to run ./run_toolbox.py instead.')
diff --git a/misc.py b/misc.py
index a9f43d5c6..3a07758b1 100644
--- a/misc.py
+++ b/misc.py
@@ -1,4 +1,4 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 import os
 import time
@@ -26,7 +26,7 @@ def __enter__(self):
     def __exit__(self, *args):
         if not self.quiet:
             titlestr = (' ' + self.title) if self.title else ''
-            print 'Elapsed%s: wall: %.06f, sys: %.06f' % ((titlestr,) + self.elapsed())
+            print('Elapsed%s: wall: %.06f, sys: %.06f' % ((titlestr,) + self.elapsed()))
 
 
 
@@ -46,7 +46,7 @@ def combine_dicts(dicts_tuple):
     '''Combines multiple dictionaries into one by adding a prefix to keys'''
     ret = {}
     for prefix,dictionary in dicts_tuple:
-        for key in dictionary.keys():
+        for key in list(dictionary.keys()):
             ret['%s%s' % (prefix, key)] = dictionary[key]
     return ret
 
@@ -57,6 +57,6 @@ def tsplit(string, no_empty_strings, *delimiters):
     pattern = '|'.join(map(re.escape, delimiters))
     strings = re.split(pattern, string)
     if no_empty_strings:
-        strings = filter(None, strings)
+        strings = [_f for _f in strings if _f]
 
-    return strings
\ No newline at end of file
+    return strings
diff --git a/optimize/gradient_optimizer.py b/optimize/gradient_optimizer.py
index e170a12cf..ad9833948 100755
--- a/optimize/gradient_optimizer.py
+++ b/optimize/gradient_optimizer.py
@@ -1,9 +1,9 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 import os
 import errno
 import pickle
-import StringIO
+import io
 from pylab import *
 from scipy.ndimage.filters import gaussian_filter
 
@@ -43,7 +43,7 @@ def __init__(self, **kwargs):
 
         self.__dict__.update(default_params)
 
-        for key,val in kwargs.iteritems():
+        for key,val in kwargs.items():
             assert key in self.__dict__, 'Unknown param: %s' % key
             self.__dict__[key] = val
 
@@ -69,10 +69,10 @@ def _validate_and_normalize(self):
         self.push_unit = (self.push_channel,) + self.push_spatial
 
     def __str__(self):
-        ret = StringIO.StringIO()
-        print >>ret, 'FindParams:'
+        ret = io.StringIO()
+        print('FindParams:', file=ret)
         for key in sorted(self.__dict__.keys()):
-            print >>ret, '%30s: %s' % (key, self.__dict__[key])
+            print('%30s: %s' % (key, self.__dict__[key]), file=ret)
         return ret.getvalue()
 
 
@@ -89,7 +89,7 @@ def __init__(self):
         self.x0 = None
         self.majority_obj = None
         self.majority_xx = None
-        self.best_obj = None
+        self.best_obj = 0
         self.best_xx = None
         self.last_obj = None
         self.last_xx = None
@@ -130,14 +130,14 @@ def trim_arrays(self):
         containing first couple values; useful for saving results as a
         reasonably sized pickle file.
         '''
-        for key,val in self.__dict__.iteritems():
+        for key,val in self.__dict__.items():
             if isinstance(val, ndarray):
                 valstr = '%s array [%s, %s, ...]' % (val.shape, val.flatten()[0], val.flatten()[1])
                 self.__dict__[key] = 'Trimmed %s' % valstr
 
     def __str__(self):
-        ret = StringIO.StringIO()
-        print >>ret, 'FindResults:'
+        ret = io.StringIO()
+        print('FindResults:', file=ret)
         for key in sorted(self.__dict__.keys()):
             val = self.__dict__[key]
             if isinstance(val, list) and len(val) > 4:
@@ -146,7 +146,7 @@ def __str__(self):
                 valstr = '%s array [%s, %s, ...]' % (val.shape, val.flatten()[0], val.flatten()[1])
             else:
                 valstr = '%s' % val
-            print >>ret, '%30s: %s' % (key, valstr)
+            print('%30s: %s' % (key, valstr), file=ret)
         return ret.getvalue()
 
 
@@ -170,14 +170,14 @@ def __init__(self, net, data_mean, labels = None, label_layers = None, channel_s
     def run_optimize(self, params, prefix_template = None, brave = False, skipbig = False):
         '''All images are in Caffe format, e.g. shape (3, 227, 227) in BGR order.'''
 
-        print '\n\nStarting optimization with the following parameters:'
-        print params
+        print('\n\nStarting optimization with the following parameters:')
+        print(params)
         
         x0 = self._get_x0(params)
         xx, results = self._optimize(params, x0)
         self.save_results(params, results, prefix_template, brave = brave, skipbig = skipbig)
 
-        print results.meta_result
+        print(results.meta_result)
         
         return xx
 
@@ -213,12 +213,12 @@ def _optimize(self, params, x0):
 
         if is_conv:
             if params.push_spatial == (0,0):
-                recommended_spatial = (data_shape[2]/2, data_shape[3]/2)
-                print ('WARNING: A unit on a conv layer (%s) is being optimized, but push_spatial\n'
+                recommended_spatial = (data_shape[2]//2, data_shape[3]//2)
+                print(('WARNING: A unit on a conv layer (%s) is being optimized, but push_spatial\n'
                        'is %s, so the upper-left unit in the channel is selected. To avoid edge\n'
                        'effects, you might want to optimize a non-edge unit instead, e.g. the center\n'
                        'unit by using `--push_spatial "%s"`\n'
-                       % (params.push_layer, params.push_spatial, recommended_spatial))
+                       % (params.push_layer, params.push_spatial, recommended_spatial)))
         else:
             assert params.push_spatial == (0,0), 'For FC layers, spatial indices must be (0,0)'
         
@@ -255,18 +255,18 @@ def _optimize(self, params, x0):
             # 3. Print progress
             if ii > 0:
                 if params.lr_policy == 'progress':
-                    print '%-4d  progress predicted: %g, actual: %g' % (ii, pred_prog, obj - old_obj)
+                    print('%-4d  progress predicted: %g, actual: %g' % (ii, pred_prog, obj - old_obj))
                 else:
-                    print '%-4d  progress: %g' % (ii, obj - old_obj)
+                    print('%-4d  progress: %g' % (ii, obj - old_obj))
             else:
-                print '%d' % ii
+                print('%d' % ii)
             old_obj = obj
 
             push_label_str = ('(%s)' % push_label) if is_labeled_unit else ''
             max_label_str  = ('(%s)' % self.labels[idxmax[0]]) if is_labeled_unit else ''
-            print '     push unit: %16s with value %g %s' % (params.push_unit, acts[params.push_unit], push_label_str)
-            print '       Max idx: %16s with value %g %s' % (idxmax, valmax, max_label_str)
-            print '             X:', xx.min(), xx.max(), norm(xx)
+            print('     push unit: %16s with value %g %s' % (params.push_unit, acts[params.push_unit], push_label_str))
+            print('       Max idx: %16s with value %g %s' % (idxmax, valmax, max_label_str))
+            print('             X:', xx.min(), xx.max(), norm(xx))
 
 
             # 4. Do backward pass to get gradient
@@ -278,9 +278,9 @@ def _optimize(self, params, x0):
             backout = self.net.backward_from_layer(params.push_layer, diffs if is_conv else diffs[:,:,0,0])
 
             grad = backout['data'].copy()
-            print '          grad:', grad.min(), grad.max(), norm(grad)
+            print('          grad:', grad.min(), grad.max(), norm(grad))
             if norm(grad) == 0:
-                print 'Grad exactly 0, failed'
+                print('Grad exactly 0, failed')
                 results.meta_result = 'Metaresult: grad 0 failure'
                 break
 
@@ -292,13 +292,13 @@ def _optimize(self, params, x0):
                 desired_prog = min(params.lr_params['early_prog'], late_prog)
                 prog_lr = desired_prog / norm(grad)**2
                 lr = min(params.lr_params['max_lr'], prog_lr)
-                print '    desired progress:', desired_prog, 'prog_lr:', prog_lr, 'lr:', lr
+                print('    desired progress:', desired_prog, 'prog_lr:', prog_lr, 'lr:', lr)
                 pred_prog = lr * dot(grad.flatten(), grad.flatten())
             elif params.lr_policy == 'progress':
                 # straight progress-based lr
                 prog_lr = params.lr_params['desired_prog'] / norm(grad)**2
                 lr = min(params.lr_params['max_lr'], prog_lr)
-                print '    desired progress:', params.lr_params['desired_prog'], 'prog_lr:', prog_lr, 'lr:', lr
+                print('    desired progress:', params.lr_params['desired_prog'], 'prog_lr:', prog_lr, 'lr:', lr)
                 pred_prog = lr * dot(grad.flatten(), grad.flatten())
             elif params.lr_policy == 'constant':
                 # constant fixed learning rate
@@ -315,7 +315,7 @@ def _optimize(self, params, x0):
 
                 if params.blur_every is not 0 and params.blur_radius > 0:
                     if params.blur_radius < .3:
-                        print 'Warning: blur-radius of .3 or less works very poorly'
+                        print('Warning: blur-radius of .3 or less works very poorly')
                         #raise Exception('blur-radius of .3 or less works very poorly')
                     if ii % params.blur_every == 0:
                         for channel in range(3):
@@ -386,12 +386,12 @@ def save_results(self, params, results, prefix_template, brave = False, skipbig
             saveimagesc('%sbest_Xpm.jpg' % prefix, asimg + self._data_mean_rgb_img)  # PlusMean
 
         with open('%sinfo.txt' % prefix, 'w') as ff:
-            print >>ff, params
-            print >>ff
-            print >>ff, results
+            print(params, file=ff)
+            print(file=ff)
+            print(results, file=ff)
         if not skipbig:
-            with open('%sinfo_big.pkl' % prefix, 'w') as ff:
+            with open('%sinfo_big.pkl' % prefix, 'wb') as ff:
                 pickle.dump((params, results), ff, protocol=-1)
         results.trim_arrays()
-        with open('%sinfo.pkl' % prefix, 'w') as ff:
+        with open('%sinfo.pkl' % prefix, 'wb') as ff:
             pickle.dump((params, results), ff, protocol=-1)
diff --git a/optimize_image.py b/optimize_image.py
index a555d9e02..477a5a611 100755
--- a/optimize_image.py
+++ b/optimize_image.py
@@ -1,4 +1,4 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 import os
 import sys
@@ -147,22 +147,22 @@ def main():
     # Load mean
     data_mean = eval(args.mean)
 
-    if isinstance(data_mean, basestring):
+    if isinstance(data_mean, str):
         # If the mean is given as a filename, load the file
         try:
             data_mean = np.load(data_mean)
         except IOError:
-            print '\n\nCound not load mean file:', data_mean
-            print 'To fetch a default model and mean file, use:\n'
-            print '  $ cd models/caffenet-yos/'
-            print '  $ cp ./fetch.sh\n\n'
-            print 'Or to use your own mean, change caffevis_data_mean in settings_local.py or override by running with `--mean MEAN_FILE` (see --help).\n'
+            print('\n\nCound not load mean file:', data_mean)
+            print('To fetch a default model and mean file, use:\n')
+            print('  $ cd models/caffenet-yos/')
+            print('  $ cp ./fetch.sh\n\n')
+            print('Or to use your own mean, change caffevis_data_mean in settings_local.py or override by running with `--mean MEAN_FILE` (see --help).\n')
             raise
         # Crop center region (e.g. 227x227) if mean is larger (e.g. 256x256)
         excess_h = data_mean.shape[1] - data_size[0]
         excess_w = data_mean.shape[2] - data_size[1]
         assert excess_h >= 0 and excess_w >= 0, 'mean should be at least as large as %s' % repr(data_size)
-        data_mean = data_mean[:, (excess_h/2):(excess_h/2+data_size[0]), (excess_w/2):(excess_w/2+data_size[1])]
+        data_mean = data_mean[:, (excess_h//2):(excess_h//2+data_size[0]), (excess_w//2):(excess_w//2+data_size[1])]
     elif data_mean is None:
         pass
     else:
@@ -172,7 +172,7 @@ def main():
         while len(data_mean.shape) < 3:
             data_mean = np.expand_dims(data_mean, -1)
 
-    print 'Using mean:', repr(data_mean)
+    print('Using mean:', repr(data_mean))
             
     # Load network
     sys.path.insert(0, os.path.join(args.caffe_root, 'python'))
diff --git a/run_toolbox.py b/run_toolbox.py
index b342fd4ae..f79351e7b 100755
--- a/run_toolbox.py
+++ b/run_toolbox.py
@@ -1,17 +1,18 @@
-#! /usr/bin/env python
+#! /usr/bin/env python3
 
 import os
 from live_vis import LiveVis
 from bindings import bindings
 try:
+#    from . import settings
     import settings
 except:
-    print '\nError importing settings.py. Check the error message below for more information.'
-    print "If you haven't already, you'll want to copy one of the settings_local.template-*.py files"
-    print 'to settings_local.py and edit it to point to your caffe checkout. E.g. via:'
-    print
-    print '  $ cp models/caffenet-yos/settings_local.template-caffenet-yos.py settings_local.py'
-    print '  $ < edit settings_local.py >\n'
+    print('\nError importing settings.py. Check the error message below for more information.')
+    print("If you haven't already, you'll want to copy one of the settings_local.template-*.py files")
+    print('to settings_local.py and edit it to point to your caffe checkout. E.g. via:')
+    print()
+    print('  $ cp models/caffenet-yos/settings_local.template-caffenet-yos.py settings_local.py')
+    print('  $ < edit settings_local.py >\n')
     raise
 
 if not os.path.exists(settings.caffevis_caffe_root):
@@ -24,7 +25,7 @@ def main():
 
     help_keys, _ = bindings.get_key_help('help_mode')
     quit_keys, _ = bindings.get_key_help('quit')
-    print '\n\nRunning toolbox. Push %s for help or %s to quit.\n\n' % (help_keys[0], quit_keys[0])
+    print('\n\nRunning toolbox. Push %s for help or %s to quit.\n\n' % (help_keys[0], quit_keys[0]))
     lv.run_loop()
 
 
diff --git a/settings.py b/settings.py
index 69c8a51bb..9c8f2ab30 100644
--- a/settings.py
+++ b/settings.py
@@ -95,9 +95,9 @@
 # All window configuation information is now contained in the
 # window_panes variable. Print if desired:
 if debug_window_panes:
-    print 'Final window panes and locations/sizes (i, j, i_size, j_size):'
+    print('Final window panes and locations/sizes (i, j, i_size, j_size):')
     for pane in window_panes:
-        print '  Pane: %s' % repr(pane)
+        print('  Pane: %s' % repr(pane))
 
 help_pane_loc = locals().get('help_pane_loc', (.07, .07, .86, .86))    # as a fraction of main window
 window_background = locals().get('window_background', (.2, .2, .2))
@@ -215,11 +215,11 @@
     caffevis_deploy_prototxt = caffevis_deploy_prototxt.replace('%DVT_ROOT%', dvt_root)
 if 'caffevis_network_weights' in locals():
     caffevis_network_weights = caffevis_network_weights.replace('%DVT_ROOT%', dvt_root)
-if isinstance(caffevis_data_mean, basestring):
+if isinstance(caffevis_data_mean, str):
     caffevis_data_mean = caffevis_data_mean.replace('%DVT_ROOT%', dvt_root)
-if isinstance(caffevis_labels, basestring):
+if isinstance(caffevis_labels, str):
     caffevis_labels = caffevis_labels.replace('%DVT_ROOT%', dvt_root)
-if isinstance(caffevis_unit_jpg_dir, basestring):
+if isinstance(caffevis_unit_jpg_dir, str):
     caffevis_unit_jpg_dir = caffevis_unit_jpg_dir.replace('%DVT_ROOT%', dvt_root)
 
 # Pause Caffe forward/backward computation for this many seconds after a keypress. This is to keep the processor free for a brief period after a keypress, which allow the interface to feel much more responsive. After this period has passed, Caffe resumes computation, in CPU mode often occupying all cores. Default: .1
diff --git a/test_keys.py b/test_keys.py
index abbbebc29..c3c406e31 100755
--- a/test_keys.py
+++ b/test_keys.py
@@ -19,30 +19,30 @@
 
 import sys
 import cv2
-import keys
-from bindings import bindings
+from . import keys
+from .bindings import bindings
 img = cv2.imread('input_images/ILSVRC2012_val_00000610.jpg')    # load example image
 
 
 
 def check_key(key_str):
-    print '  Press key %5s: ' % key_str,
+    print('  Press key %5s: ' % key_str, end=' ')
     sys.stdout.flush()
     while True:
         keycode = cv2.waitKey(0)
         label, masked_vals = bindings.get_key_label_from_keycode(keycode, extra_info = True)
         if label and ('shift' in label or 'ctrl' in label):
-            print '(ignoring modifier %s)' % label,
+            print('(ignoring modifier %s)' % label, end=' ')
             sys.stdout.flush()
         else:
             break
     masked_vals_pp = ', '.join(['%d (%s)' % (mv, hex(mv)) for mv in masked_vals])
     if label == key_str:
-        print '  %d (%s) matched %s' % (keycode, hex(keycode), label)
+        print('  %d (%s) matched %s' % (keycode, hex(keycode), label))
     elif label is not None:
-        print '* %d (%s) failed, matched key %s (masked vals tried: %s)' % (keycode, hex(keycode), label, masked_vals_pp)
+        print('* %d (%s) failed, matched key %s (masked vals tried: %s)' % (keycode, hex(keycode), label, masked_vals_pp))
     else:
-        print '* %d (%s) failed, no match found (masked vals tried: %s)' % (keycode, hex(keycode), masked_vals_pp)
+        print('* %d (%s) failed, no match found (masked vals tried: %s)' % (keycode, hex(keycode), masked_vals_pp))
     #print 'Got:', label
     #found = False
     #for k,v in keys.Key.__dict__.iteritems():
@@ -58,7 +58,7 @@ def check_key(key_str):
 
     
 def main():
-    print 'Click on the picture and then carefully push the following keys:'
+    print('Click on the picture and then carefully push the following keys:')
     cv2.imshow('img',img)
     check_key('j')
     check_key('k')