diff --git a/README.md b/README.md
index 92c5604..6d8333c 100644
--- a/README.md
+++ b/README.md
@@ -78,7 +78,11 @@ Step by step:
 
   1. Open the `main/` directory
   2. Insert the input images and videos in the folder **input/**
-  3. Insert the classes in the file **class_list.txt** (one class name per line)
+  3. Edit config.in setting the full filepath to your class_list (WITHOUT quotation marks). E.g.
+     [CLASSES]
+     MOST_RECENT_FILE = /home/user1/Desktop/class_list.txt
+     
+     Alternatively, just edit the contents of the example file ./main/class_list.txt
   4. Run the code:
   5. You can find the annotations in the folder **output/**
 
diff --git a/main/__init__.py b/main/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/main/config.ini b/main/config.ini
index fb305f9..a9b54ce 100644
--- a/main/config.ini
+++ b/main/config.ini
@@ -2,3 +2,6 @@
 OBJECT_SCORE_THRESHOLD = 0.65
 OBJECT_IDS = 1,2
 CUDA_VISIBLE_DEVICES = ''
+
+[CLASSES]
+MOST_RECENT_FILE =
diff --git a/main/load_classes.py b/main/load_classes.py
new file mode 100644
index 0000000..b66536a
--- /dev/null
+++ b/main/load_classes.py
@@ -0,0 +1,32 @@
+import configparser
+
+from pathlib import Path
+
+
+def non_blank_lines(file_object):
+    for l in file_object:
+        line = l.rstrip()
+        if line:
+            yield line
+
+
+def get_class_list():
+    """
+    Uses the most recent classes source file defined in config.ini
+    otherwise defaults the example class_list.txt file provided in
+    this repository.
+
+    """
+    config = configparser.ConfigParser()
+    this_dir = Path(__file__).parent
+    config_path = this_dir / "config.ini"
+    config.read(str(config_path))
+    most_recent_classes_file = config.get("CLASSES", "MOST_RECENT_FILE")
+
+    if not most_recent_classes_file:
+        classes_src = str(this_dir / "class_list.txt")
+    else:
+        classes_src = most_recent_classes_file
+
+    with open(classes_src) as f:
+        return list(non_blank_lines(file_object=f))
diff --git a/main/main.py b/main/main.py
index 0b76409..05da04c 100755
--- a/main/main.py
+++ b/main/main.py
@@ -1,6 +1,5 @@
 #!/bin/python
 import argparse
-import glob
 import json
 import os
 import re
@@ -12,7 +11,10 @@
 from lxml import etree
 import xml.etree.cElementTree as ET
 
+from load_classes import get_class_list
 
+
+CLASS_LIST = get_class_list()
 DELAY = 20 # keyboard delay (in milliseconds)
 WITH_QT = False
 try:
@@ -718,13 +720,6 @@ def convert_video_to_images(video_path, n_frames, desired_img_format):
     return file_path, video_name_ext
 
 
-def nonblank_lines(f):
-    for l in f:
-        line = l.rstrip()
-        if line:
-            yield line
-
-
 def get_annotation_paths(img_path, annotation_formats):
     annotation_paths = []
     for ann_dir, ann_ext in annotation_formats.items():
@@ -1032,10 +1027,6 @@ def complement_bgr(color):
                 elif '.xml' in ann_path:
                     create_PASCAL_VOC_xml(ann_path, abs_path, folder_name, image_name, img_height, img_width, depth)
 
-    # load class list
-    with open('class_list.txt') as f:
-        CLASS_LIST = list(nonblank_lines(f))
-    #print(CLASS_LIST)
     last_class_index = len(CLASS_LIST) - 1
 
     # Make the class colors the same each session
diff --git a/main/main_auto.py b/main/main_auto.py
index b32e09f..b8be738 100755
--- a/main/main_auto.py
+++ b/main/main_auto.py
@@ -26,6 +26,12 @@
 from lxml import etree
 import xml.etree.cElementTree as ET
 import sys
+
+from load_classes import get_class_list
+
+
+CLASS_LIST = get_class_list()
+
 sys.path.insert(0, "..")
 from object_detection.tf_object_detection import ObjectDetector
 import configparser
@@ -313,13 +319,6 @@ def convert_video_to_images(video_path, n_frames, desired_img_format):
     return file_path, video_name_ext
 
 
-def nonblank_lines(f):
-    for l in f:
-        line = l.rstrip()
-        if line:
-            yield line
-
-
 def get_annotation_paths(img_path, annotation_formats):
     annotation_paths = []
     for ann_dir, ann_ext in annotation_formats.items():
@@ -695,10 +694,6 @@ def predict_next_frames(self,json_file_data,json_file_path):
             elif '.xml' in ann_path:
                 create_PASCAL_VOC_xml(ann_path, abs_path, folder_name, image_name, img_height, img_width, depth)
 
-# load class list
-with open('class_list.txt') as f:
-    CLASS_LIST = list(nonblank_lines(f))
-#print(CLASS_LIST)
 last_class_index = len(CLASS_LIST) - 1
 
 # Make the class colors the same each session
diff --git a/requirements.txt b/requirements.txt
index db49ed2..8286b81 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,4 +1,4 @@
-lxml==4.3.0
-numpy==1.16.0
+lxml==4.6.3
+numpy==1.21.2
 opencv-contrib-python==3.4.9.33
 tqdm==4.29.1
diff --git a/tests/__init__.py b/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/test_load_classes.py b/tests/test_load_classes.py
new file mode 100644
index 0000000..9df76d7
--- /dev/null
+++ b/tests/test_load_classes.py
@@ -0,0 +1,6 @@
+from main.load_classes import get_class_list
+
+
+def test_get_class_list():
+    classes = get_class_list()
+    assert classes == ['person', 'billiard ball', 'donut']