Added comments to evaluation scripts

benchmark_testing/README.md

@@ -40,6 +40,7 @@ Note: One such pair of sample database and queries directories have already been
 
     > [!NOTE]\
     > *All the code below should be executed from the `benchmark_testing` directory.*
+    > Make sure to execute all .sh files using git bash.
 
 2. **Run Test Data Setup (Optional)**
    - Replace the directory names in the `test_data_setup.py` file with appropriate directory locations.

benchmark_testing/run_bulk_upload.sh

@@ -13,8 +13,11 @@ sleep 10
 # Start timer
 start_time=$(date +%s)
 
-# Run your functions or any additional Python commands here
-python ../src/Sample_Client/sample_bulk_upload_client.py --directory_paths "\path\to\input\directory" --database_name "sample_db"
+# Call client script to upload images from to database (the code currently only accepts one directory at a time)
+python ../src/Sample_Client/sample_bulk_upload_client.py --directory_paths "\\path\\to\\sample_database\\directory" --database_name "sample_db"
+
+# Sample directory path
+# "<path to dataset folder>\\LFWdataset\\sample_database"
 
 # Calculate total time taken
 end_time=$(date +%s)

benchmark_testing/run_face_find_accuracy.sh

@@ -14,8 +14,12 @@ sleep 10
 start_time=$(date +%s)
 
 # Define the directory containing the files and the output CSV file
-input_directory="\path\to\input\directory"
-output_csv="\path\to\output\csv"
+input_directory="\\path\\to\\input\\directory" # Path to directory of query images
+output_csv="\\path\\to\\output\\csv" # Path to csv file containing top-n matches
+
+# Sample directory path
+# input_directory = "<path to dataset folder>\\LFWdataset\\sample_queries"
+# output_csv = "<path to dataset folder>\\LFWdataset\\output.csv"
 
 # Initialize the CSV file with headers
 echo "filename,result" > "$output_csv"

benchmark_testing/run_face_find_time.sh

@@ -13,9 +13,12 @@ sleep 10
 # Start timer
 start_time=$(date +%s)
 
-# Run your functions or any additional Python commands here
+# Call client script to find match for an image (code currently only accepts one file_path at a time)
 python ../src/Sample_Client/sample_find_face_client.py --file_paths "\path\to\image" --database_name "sample_db"
 
+# Sample file path
+# "<path to dataset folder>\\LFWdataset\\sample_queries\\image.jpg"
+
 # Calculate total time taken
 end_time=$(date +%s)
 total_time=$((end_time - start_time))

benchmark_testing/test_data_metrics.py

@@ -4,7 +4,11 @@
 from sklearn.metrics import accuracy_score
 
 # Load the CSV file
-file_path = "path\\to\\output\\csv"
+file_path = "path\\to\\output\\csv"  # Path to csv file containing top-n matches
+
+# Sample csv path
+# file_path = "<path to dataset folder>\\LFWdataset\\output.csv"
+
 data = pd.read_csv(file_path)
 
 # Extract ground truth names (base names without numeric suffixes)

benchmark_testing/test_data_setup.py

@@ -47,10 +47,21 @@ def count_files(directory):
     )
 
 
-# Specify the root directory you want to start from
-root_directory = "\\path\\to\\root\\directory"
-upload_directory = "\\path\\to\\upload\\directory"
-test_directory = "\\path\\to\\test\\directory"
+# Edit the following paths to match your system's directories
+
+root_directory = "\\path\\to\\root\\directory"  # Original dataset directory
+upload_directory = "\\path\\to\\upload\\directory"  # Directory where images to be uploaded to database will be stored
+test_directory = (
+    "\\path\\to\\test\\directory"  # Directory where query images will be stored
+)
+
+"""
+Sample directories
+root_directory = "<path to dataset folder>\\LFWdataset\\lfw"
+upload_directory = "<path to dataset folder>\\LFWdataset\\sample_database" 
+test_directory = "<path to dataset folder>\\LFWdataset\\sample_queries" 
+"""
+
 process_directory(root_directory, upload_directory, test_directory)
 
 # Print the number of files in the test and verify directories