A system for identifying latent scene attributes in videos (vlsa) using knowledge acquired from large collections of text documents.

This project is maintained by trananh



More information coming soon ...


Zip Downloads

Youtube Playlists

Installation & Compilation

  1. First, clone/download the project to a local repository.
  2. In the terminal, change directory to the root of the project and run this line once to install external jars.

Command-line Compilation

Then, every time you need to compile the project:

mvn compile

Test the project using:

mvn test

Using IntelliJ IDEA

Install IntelliJ IDEA

If you have not done so already, download IntelliJ IDEA Community Edition from here.

Install the Scala plugin. See this page for installation details

Import vlsa

When you are done installing IntelliJ, you are ready to import vlsa. Go to File/Import Project. In the file window that opens select vlsa/pom.xml. In the next window, make sure that only the following options are checked:

Search for projects recursively
Project format: .idea (directory based)
Import Maven projects automatically
Create IDEA modules for aggregator projects (with 'pom' packaging)
Create module groups for multi-module Maven projects
Keep source and test folders in reimport
Exclude build directory (%PROJECT_ROOT%/target)
Use Maven output directories
Generated source folders: Detect automatically
Phase to be used for folders update: process-resources
Automatically download: Sources (YES) Documentation (YES)

Click Next. In the second window, check the Maven project edu.arizona.sista:vlsa. Click Next. In the next window select SDK 1.6 (should be already selected). Click Next. In the final window, make sure the project name is vlsa. You can also choose the project directory location on your disk here. Click Finish. The result should be an IntelliJ project called vlsa.

Before using vlsa, please do the following:

Compile and test

After this is done, compile the project. It should compile without errors (maybe a few warnings). If you see this warning:

Unregistered Git root detected

Click the popup to go to the configuration window and accept the proposed configuration with the new Git root (click Ok).

Once the project has compiled run a test program to double check. For example this:


To run, right click on the file, and choose the Run option. You may need to give the programs sufficient memory to run. To adjust find Edit Configurations, then add -Xmx3g in the VM Options field.