Update README.md
Browse files
README.md
CHANGED
|
@@ -3,18 +3,12 @@ language: "multilingual"
|
|
| 3 |
license: "mit"
|
| 4 |
---
|
| 5 |
|
| 6 |
-
Unidrone v1.0
|
| 7 |
-
|
| 8 |
-
---------------------------------------------------------------------
|
| 9 |
|
| 10 |
Welcome to the Unidrone v1.0 AI models 🚁🎉
|
| 11 |
|
| 12 |
-
---------------------------------------------------------------------
|
| 13 |
|
| 14 |
What is Unidrone?
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|
| 18 |
Unidrone is a a set of detection AI models, based on a large YOLO-v8 backbone and my own data pipeline. It's mostly a merging of two of my other models: WALDO and NANO. The former is great at detections in nadir imagery (straight overhead) and the latter is better for "forward-looking" angles. Before you had to choose one of those models and stick to a limited set of angles, but now Unidrone does both really well!
|
| 19 |
|
| 20 |
The AI models are capable of detecting these classes of items in drone imagery ranging in altitude from about 50 feet to 1000 feet, regardless of angle:
|
|
|
|
| 3 |
license: "mit"
|
| 4 |
---
|
| 5 |
|
|
|
|
|
|
|
|
|
|
| 6 |
|
| 7 |
Welcome to the Unidrone v1.0 AI models 🚁🎉
|
| 8 |
|
|
|
|
| 9 |
|
| 10 |
What is Unidrone?
|
| 11 |
|
|
|
|
|
|
|
| 12 |
Unidrone is a a set of detection AI models, based on a large YOLO-v8 backbone and my own data pipeline. It's mostly a merging of two of my other models: WALDO and NANO. The former is great at detections in nadir imagery (straight overhead) and the latter is better for "forward-looking" angles. Before you had to choose one of those models and stick to a limited set of angles, but now Unidrone does both really well!
|
| 13 |
|
| 14 |
The AI models are capable of detecting these classes of items in drone imagery ranging in altitude from about 50 feet to 1000 feet, regardless of angle:
|