don't use jet as default.

examples/ConvolutionalPoseMachines/load-cpm.py

@@ -11,9 +11,6 @@ import argparse
 from tensorpack import *
 from tensorpack.utils.argtools import memoized
 
-import matplotlib.pyplot as plt
-_CM = plt.get_cmap('jet')
-
 """
 15 channels:
 0-1 head, neck
@@ -29,7 +26,7 @@ def colorize(img, heatmap):
     """ img: bgr, [0,255]
         heatmap: [0,1]
     """
-    heatmap = _CM(heatmap)[:, :, [2, 1, 0]] * 255.0
+    heatmap = viz.intensity_to_rgb(heatmap, cmap='jet')[:, :, ::-1]
     return img * 0.5 + heatmap * 0.5
 
 

examples/SpatialTransformer/README.md

@@ -9,7 +9,7 @@ produce the sum of them.
 Here the two Spatial Transformer branches learn to localize the two digits
 and warped them separately.
 
-![demo](demo.jpg)
+<p align="center"> <img src="./demo.jpg" width="400"> </p>
 
 Left: input image; Middle: output of the first STN branch (which localizes the second digit); Right: output of the second STN branch.
 

examples/saliency/saliency-maps.py

@@ -2,13 +2,13 @@
 # -*- coding: utf-8 -*-
 
 import cv2
-import cPickle as pickle
 import sys
 import os
 import numpy as np
 import tensorflow as tf
 import tensorflow.contrib.slim as slim
 from tensorflow.contrib.slim.nets import resnet_v1
+
 import tensorpack as tp
 import tensorpack.utils.viz as viz
 

tensorpack/train/base.py

@@ -102,8 +102,11 @@ class Trainer(object):
         add scalar summary to ``self.stat_holder``.
 
         Args:
-            summary (tf.Summary): a summary object.
+            summary (tf.Summary or str): a summary object, or a str which will
+                be interpreted as a serialized tf.Summary protobuf.
         """
+        if isinstance(summary, six.string_types):
+            summary = tf.Summary.FromString(summary)
         for val in summary.value:
             if val.WhichOneof('value') == 'simple_value':
                 val.tag = re.sub('tower[p0-9]+/', '', val.tag)   # TODO move to subclasses
@@ -116,10 +119,6 @@ class Trainer(object):
     def add_scalar_summary(self, name, val):
         """
         Add a scalar sumary to both TF events file and StatHolder.
-
-        Args:
-            name(str)
-            val(float)
         """
         self.add_summary(create_scalar_summary(name, val))
 

tensorpack/train/feedfree.py

@@ -29,7 +29,7 @@ class FeedfreeTrainerBase(Trainer):
         # note that summary_op will take a data from the queue
         if self.summary_op is not None:
             summary_str = self.summary_op.eval()
-            self.add_summary(tf.Summary.FromString(summary_str))
+            self.add_summary(summary_str)
 
     def _get_input_tensors(self):
         return self._input_method.get_input_tensors()

tensorpack/train/trainer.py

@@ -95,7 +95,7 @@ class SimpleTrainer(Trainer):
         if self.summary_op is not None:
             feed = self._input_method.next_feed()
             summary_str = self.summary_op.eval(feed_dict=feed)
-            self.add_summary(tf.Summary.FromString(summary_str))
+            self.add_summary(summary_str)
 
     def get_predict_func(self, input_names, output_names):
         return self._predictor_factory.get_predictor(input_names, output_names, 0)

tensorpack/utils/viz.py

@@ -243,7 +243,7 @@ def dump_dataflow_images(df, index=0, batched=True,
                 vizlist = vizlist[vizsize:]
 
 
-def intensity_to_rgb(intensity, cmap='jet', normalize=False):
+def intensity_to_rgb(intensity, cmap='cubehelix', normalize=False):
     """
     Convert a 1-channel matrix of intensities to an RGB image employing a colormap.
     This function requires matplotlib. See `matplotlib colormaps