A prototype implementation of VOCAL-UDF, which is a self-enhancing video data management system that empowers users to flexibly issue and answer compositional queries, even when the modules necessary to answer those queries are unavailable. See the technical report for more details.
The project uses conda to manage dependencies. To install conda, follow the instructions here.
# Clone the repository
git clone https://github.com/uwdb/VOCAL-UDF.git
cd VOCAL-UDF
# Create a conda environment (called vocal-udf) and install dependencies
conda env create -f environment.yml --name vocal-udf
conda activate vocal-udf
python -m pip install -e .TODO
cd scripts- Generate UDFs
./exp-vocal_udf_main_{dataset}.sh- Execute query with new UDFs
./exp-vocal_udf_query_execution_{dataset}.sh