Spaces:
No application file
No application file
Update README.md
Browse files
README.md
CHANGED
|
@@ -9,4 +9,21 @@ app_file: app.py
|
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
|
| 12 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 9 |
pinned: false
|
| 10 |
---
|
| 11 |
|
| 12 |
+
# UAV Drone Detection Frames
|
| 13 |
+
|
| 14 |
+
This dataset contains frames with at least one drone detection, stored in Parquet format.
|
| 15 |
+
|
| 16 |
+
## Features
|
| 17 |
+
|
| 18 |
+
- `video`: source video name
|
| 19 |
+
- `frame_file`: frame filename
|
| 20 |
+
- `image`: embedded image
|
| 21 |
+
- `objects.bbox`: bounding boxes in `[x, y, width, height]`
|
| 22 |
+
- `objects.categories`: class ids
|
| 23 |
+
- `scores`: detection confidence scores
|
| 24 |
+
|
| 25 |
+
## Description
|
| 26 |
+
|
| 27 |
+
The dataset was created by extracting frames from drone videos, running a fine-tuned YOLO detector, and keeping only frames with at least one detection.
|
| 28 |
+
|
| 29 |
+
This dataset was prepared for an academic drone detection and tracking assignment at NYU for the Intro to AI class Spring 2026.
|