diff --git a/README.md b/README.md
index dc22663..d46e781 100755
--- a/README.md
+++ b/README.md
@@ -5,7 +5,7 @@
 
 Welcome to the [Ultralytics xView YOLOv3](https://github.com/ultralytics/xview-yolov3) repository! Here we provide code to train the powerful YOLOv3 object detection model on the xView dataset for the [xView Challenge](https://challenge.xviewdataset.org/). This challenge focuses on detecting objects from satellite imagery, advancing the state of the art in computer vision applications for remote sensing.
 
-[![Ultralytics Actions](https://github.com/ultralytics/xview-yolov3/actions/workflows/format.yml/badge.svg)](https://github.com/ultralytics/xview-yolov3/actions/workflows/format.yml) <a href="https://ultralytics.com/discord"><img alt="Discord" src="https://img.shields.io/discord/1089800235347353640?logo=discord&logoColor=white&label=Discord&color=blue"></a>
+[![Ultralytics Actions](https://github.com/ultralytics/xview-yolov3/actions/workflows/format.yml/badge.svg)](https://github.com/ultralytics/xview-yolov3/actions/workflows/format.yml) <a href="https://ultralytics.com/discord"><img alt="Discord" src="https://img.shields.io/discord/1089800235347353640?logo=discord&logoColor=white&label=Discord&color=blue"></a> <a href="https://community.ultralytics.com"><img alt="Ultralytics Forums" src="https://img.shields.io/discourse/users?server=https%3A%2F%2Fcommunity.ultralytics.com&logo=discourse&label=Forums&color=blue"></a>
 
 <img src="https://github-production-user-asset-6210df.s3.amazonaws.com/26833433/238799379-bb3b02f0-dee4-4e67-80ae-4b2378b813ad.jpg?raw=true" width="100%">
 
diff --git a/utils/datasets.py b/utils/datasets.py
index 09b90bd..06cd990 100755
--- a/utils/datasets.py
+++ b/utils/datasets.py
@@ -301,6 +301,9 @@ def resize_square(img, height=416, color=(0, 0, 0)):  # resizes a rectangular im
 def random_affine(
     img, targets=None, degrees=(-10, 10), translate=(0.1, 0.1), scale=(0.9, 1.1), shear=(-3, 3), borderValue=(0, 0, 0)
 ):
+    """Performs random affine transformations on an image and its target annotations, including rotation, translation,
+    scaling, and shearing.
+    """
     # torchvision.transforms.RandomAffine(degrees=(-10, 10), translate=(.1, .1), scale=(.9, 1.1), shear=(-10, 10))
     # https://medium.com/uruvideo/dataset-augmentation-with-random-homographies-a8f4b44830d4
     border = 750
diff --git a/utils/utils.py b/utils/utils.py
index 17e0969..b2dbebd 100755
--- a/utils/utils.py
+++ b/utils/utils.py
@@ -517,15 +517,9 @@ def build_targets(pred_boxes, pred_conf, pred_cls, target, anchor_wh, nA, nC, nG
 
 
 def non_max_suppression(prediction, conf_thres=0.5, nms_thres=0.4, mat=None, img=None, model2=None, device="cpu"):
-    prediction = prediction.cpu()
-    """
-    Removes detections with lower object confidence score than 'conf_thres' and performs Non-Maximum Suppression to
-    further filter detections.
-
-    Returns detections with shape:
-        (x1, y1, x2, y2, object_conf, class_score, class_pred)
-    """
+    """Performs Non-Maximum Suppression on detection results, filtering out low-confidence detections."""
 
+    prediction = prediction.cpu()
     output = [None for _ in range(len(prediction))]
     # Gather bbox priors
     srl = 6  # sigma rejection level