Upload README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,30 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
# Pistol Detection
|
| 2 |
+
Detecting Pistol in the Video / Photo
|
| 3 |
+
|
| 4 |
+
## Output
|
| 5 |
+
|
| 6 |
+
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
## Installation
|
| 10 |
+
|
| 11 |
+
##### SETUP
|
| 12 |
+
```shell
|
| 13 |
+
#---------Using MAKE (Recommended) ----------
|
| 14 |
+
make provision
|
| 15 |
+
conda activate aero-detection
|
| 16 |
+
make build
|
| 17 |
+
```
|
| 18 |
+
|
| 19 |
+
Trained 250 Epochs
|
| 20 |
+
|
| 21 |
+
The dataset includes 2971 images.
|
| 22 |
+
Pistol are annotated in YOLOv8 format.
|
| 23 |
+
|
| 24 |
+
The following pre-processing was applied to each image:
|
| 25 |
+
* Auto-orientation of pixel data (with EXIF-orientation stripping)
|
| 26 |
+
* Resize to 640x640 (Stretch)
|
| 27 |
+
|
| 28 |
+
No image augmentation techniques were applied.
|
| 29 |
+
|
| 30 |
+
Dataset Link : https://www.kaggle.com/datasets/cpluzshrijayan
|