add notes

_posts/deep_learning/2015-10-09-dl-applications.md

@@ -789,6 +789,13 @@ Illustration Tagger, InceptionNet, Parsey McParseface, ArtsyNetworks
 - intro: Extentions of mmm 2017 paper
 - arxiv: [https://arxiv.org/abs/1703.08173](https://arxiv.org/abs/1703.08173)
 
+## Video Super-resolution
+
+**Detail-revealing Deep Video Super-resolution**
+
+- arxiv: [https://arxiv.org/abs/1704.02738](https://arxiv.org/abs/1704.02738)
+- github: [https://github.com/jiangsutx/SPMC_VideoSR](https://github.com/jiangsutx/SPMC_VideoSR)
+
 # Image Denoising
 
 **Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising**

_posts/deep_learning/2015-10-09-image-retrieval.md

@@ -336,6 +336,10 @@ and from 73.8% to 83.1% for ResNet-50
 - intro: ACCV 2016
 - arxiv: [https://arxiv.org/abs/1612.03900](https://arxiv.org/abs/1612.03900)
 
+**Supervised Deep Hashing for Hierarchical Labeled Data**
+
+- arxiv: [https://arxiv.org/abs/1704.02088](https://arxiv.org/abs/1704.02088)
+
 # Learning to Rank
 
 **Simple to Complex Cross-modal Learning to Rank**

_posts/deep_learning/2015-10-09-segmentation.md

@@ -445,6 +445,11 @@ with built-in awareness of semantically meaningful boundaries. "
 - intro: CVPR 2017 spotlight paper
 - arxxiv: [https://arxiv.org/abs/1704.01344](https://arxiv.org/abs/1704.01344)
 
+**Loss Max-Pooling for Semantic Image Segmentation**
+
+- intro: CVPR 2017
+- arxiv: [https://arxiv.org/abs/1704.02966](https://arxiv.org/abs/1704.02966)
+
 # Instance Segmentation
 
 **Simultaneous Detection and Segmentation**
@@ -592,6 +597,11 @@ with built-in awareness of semantically meaningful boundaries. "
 
 [https://arxiv.org/abs/1704.01152](https://arxiv.org/abs/1704.01152)
 
+**Pixelwise Instance Segmentation with a Dynamically Instantiated Network**
+
+- intro: CVPR 2017
+- arxiv: [https://arxiv.org/abs/1704.02386](https://arxiv.org/abs/1704.02386)
+
 # Specific Segmentation
 
 **End-to-end semantic face segmentation with conditional random fields as convolutional, recurrent and adversarial networks**

drafts/2017-04-13-notes-on-tensorflow-dev.md

@@ -0,0 +1,111 @@
+---
+layout: post
+category: deep_learning
+title: Notes On Tensorflow Development
+date: 2017-04-13
+---
+
+# Install Bazel on Ubuntu 14.04
+
+**Install with installer:**
+
+[https://bazel.build/versions/master/docs/install-ubuntu.html#install-with-installer](https://bazel.build/versions/master/docs/install-ubuntu.html#install-with-installer)
+
+Official instruction is simple:
+
+```
+sudo apt-get install openjdk-8-jdk
+```
+
+But due to network/proxy problem I cannot make it work. 
+After some goolings I found a work-around if someone else also fails on installing openjdk-8-jdk following above command: intalling Oracle's jdk.
+
+Step 1: Download jdk8
+
+I choose to use jdk-8u121-linux-x64.tar.gz:
+
+[http://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html](http://www.oracle.com/technetwork/java/javase/downloads/jdk8-downloads-2133151.html)
+
+Step 2: Uncompression and Copy to one target directory:
+
+```
+sudo mkdir /usr/lib/jvm
+sudo tar -zxvf jdk-8u121-linux-x64.tar.gz -C /usr/lib/jvm
+```
+
+Step 3: Modify some system PATHs:
+
+Add following lines to your ~/.bashrc :
+
+```
+#set oracle jdk environment @20170413
+export JAVA_HOME=/usr/lib/jvm/jdk1.8.0_121
+export JRE_HOME=${JAVA_HOME}/jre
+export CLASSPATH=.:${JAVA_HOME}/lib:${JRE_HOME}/lib
+export PATH=${JAVA_HOME}/bin:$PATH
+```
+
+Don't forget to enable these settings:
+
+```
+source ~/.bashrc
+```
+
+Step 4: Set system default jdk version:
+
+```
+jdk_ver=jdk1.8.0_121
+
+sudo update-alternatives --install /usr/bin/java java /usr/lib/jvm/${jdk_ver}/bin/java 300
+sudo update-alternatives --install /usr/bin/javac javac /usr/lib/jvm/${jdk_ver}/bin/javac 300
+sudo update-alternatives --install /usr/bin/jar jar /usr/lib/jvm/${jdk_ver}/bin/jar 300
+sudo update-alternatives --install /usr/bin/javah javah /usr/lib/jvm/${jdk_ver}/bin/javah 300
+sudo update-alternatives --install /usr/bin/javap javap /usr/lib/jvm/${jdk_ver}/bin/javap 300
+
+sudo update-alternatives --config java
+```
+
+You might get similar outputs like below:
+
+```
+my_account@node7:~/my_account/sw$ sudo ./set_system_default_jdk_version.sh
+update-alternatives: using /usr/lib/jvm/jdk1.8.0_121/bin/javac to provide /usr/bin/javac (javac) in auto mode
+update-alternatives: using /usr/lib/jvm/jdk1.8.0_121/bin/jar to provide /usr/bin/jar (jar) in auto mode
+update-alternatives: using /usr/lib/jvm/jdk1.8.0_121/bin/javah to provide /usr/bin/javah (javah) in auto mode
+update-alternatives: using /usr/lib/jvm/jdk1.8.0_121/bin/javap to provide /usr/bin/javap (javap) in auto mode
+There are 2 choices for the alternative java (providing /usr/bin/java).
+
+  Selection    Path                                Priority   Status
+------------------------------------------------------------
+* 0            /usr/bin/gij-4.8                     1048      auto mode
+  1            /usr/bin/gij-4.8                     1048      manual mode
+  2            /usr/lib/jvm/jdk1.8.0_121/bin/java   300       manual mode
+
+Press enter to keep the current choice[*], or type selection number: 2
+update-alternatives: using /usr/lib/jvm/jdk1.8.0_121/bin/java to provide /usr/bin/java (java) in manual mode
+```
+
+Run `java -version`, you will get following output:
+
+```
+my_account@node7:~/my_account/sw$ java -version
+java version "1.8.0_121"
+Java(TM) SE Runtime Environment (build 1.8.0_121-b13)
+Java HotSpot(TM) 64-Bit Server VM (build 25.121-b13, mixed mode)
+```
+
+Step 5:
+
+And finally you can successfully execute below command:
+
+```
+bazel-0.4.5-installer-linux-x86_64.sh --user
+```
+
+It would be useful to add following line to your ~/.bashrc :
+```
+# for Bazel
+export PATH="$PATH:$HOME/bin"
+```
+
+Done.