This is a UI testing for trained model using DeepLearning4J Library. Some part of my code and images or videos were from this link so if you want to get more knowledge about DeepLearning, you may want to check on this repo.
For this project, I have used IntelliJ IDE for faster development. Assuming at least Java JDK 1.8 and JavaFX Library are installed already and you cloned or downloaded this repository.
This is a step by step process to get this code running:
First step. Install VLC but this is optional if the version of libavcodec.so you have is compatible with the version of Linux or Ubuntu machine and JavaFX Library installed. If not then go through this steps because we're not going to use MediaPlayer from JavaFX Library because I can't get it to work with my Ubuntu version unless it's working for you:
sudo aptitude
then search for vlc and install
Second step. Add this to your build.gradle
file if you are building with gradle:
// dl4j
compile group: 'org.nd4j', name: 'nd4j-native-platform', version: '1.0.0-beta2'
compile group: 'org.deeplearning4j', name: 'deeplearning4j-core', version: '1.0.0-beta2'
compile group: 'org.deeplearning4j', name: 'deeplearning4j-nlp', version: '1.0.0-beta2'
compile group: 'org.datavec', name: 'datavec-api', version: '1.0.0-beta2'
compile group: 'org.deeplearning4j', name: 'deeplearning4j-zoo', version: '1.0.0-beta2'
// logs
compile group: 'org.slf4j', name: 'slf4j-api', version: '1.7.25'
compile group: 'org.slf4j', name: 'slf4j-log4j12', version: '1.7.25'
// bytedeco
compile group: 'org.bytedeco.javacpp-presets', name: 'ffmpeg', version: '4.0.2-1.4.3'
compile group: 'org.bytedeco.javacpp-presets', name: 'ffmpeg-platform', version: '4.0.2-1.4.3'
// vlcj
compile group: 'uk.co.caprica', name: 'vlcj', version: '3.10.1'
Third step. If models
directory is not present in resources
, go ahead and make one then add the model.zip
file to home/dir/deeplearningui/resources/models/
that you are going to use later.
Fourth step. Modify the LocalData.java
and put your customized classes if you have one.
You can choose a model from specific directory by clicking the button with folder icon at the left top corner or use the pretrained model from dl4j link
Load your Validation
or Test
dataset
Click one picture or video you want your model to start testing object detection
Useful Links