Hugging Face's logo

blackbeedrones
/
imav-2025-platform

Object Detection
ultralytics
TensorBoard
ONNX
yolo
yolov11
drone
uav
imav
robotics
autonomous-landing
helipad-detection
Eval Results (legacy)
Model card Files
xet
 Metrics Community
samuellimabraz commited on
Commit
3b0a123
·
verified ·
1 Parent(s): 59e9edf

Update README with mission details and platform specs: README.md

Browse files
Files changed (1) hide show
  1. README.md +44 -30
README.md CHANGED
@@ -6,9 +6,11 @@ tags:
6
  - yolov11
7
  - ultralytics
8
  - drone
 
9
  - imav
10
  - robotics
11
- - search-and-rescue
 
12
  datasets:
13
  - custom
14
  pipeline_tag: object-detection
@@ -16,17 +18,41 @@ pipeline_tag: object-detection
16
 
17
  # IMAV 2025 Platform Detection - YOLOv11n
18
 
19
- Platform detection model for the **IMAV 2025 Indoor Competition - Mission 4 (Platform Localization & Landing)**.
20
 
21
- ## Model Description
22
 
23
- This YOLOv11n model was trained to detect landing platforms/bases used in the IMAV 2025 indoor competition. The model is designed for real-time detection on autonomous drones.
24
 
25
- ### Competition Context
26
 
27
- The [16th International Micro Air Vehicle Conference and Competition (IMAV 2025)](https://femexrobotica.org/imav2025/) took place in San Andrés Cholula, Puebla, Mexico. The competition theme was **"Search and Rescue"**, inspired by Mexico's seismic activity and the need for micro air vehicles in disaster response scenarios.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
28
 
29
- **Mission 4**: The objective is to land on a moving platform with smoke.
 
 
 
 
 
30
 
31
  ## Model Formats
32
 
@@ -34,52 +60,40 @@ The [16th International Micro Air Vehicle Conference and Competition (IMAV 2025)
34
  |--------|------|----------|
35
  | PyTorch | `platform_yolov11n.pt` | Training, fine-tuning |
36
  | ONNX | `platform_yolov11n.onnx` | Cross-platform inference |
37
- | TensorRT | `platform_yolov11n.engine` | NVIDIA Jeston Orin Nano Super optimized inference |
38
 
39
  ## Usage
40
 
41
- ### With mirela-sdk
42
 
43
  ```python
44
  from mirela_sdk.ai.detection import Detector
45
 
46
- # Load model
47
  detector = Detector("blackbeedrones/imav-2025-platform:platform_yolov11n.pt")
48
  detector.load()
49
 
50
- # Detect platforms
51
  result = detector.detect(image, conf=0.5)
52
  for det in result:
53
- print(f"Platform detected: {det.confidence:.2f} at {det.bbox}")
54
  ```
55
 
56
- ### With Ultralytics
57
 
58
  ```python
59
  from ultralytics import YOLO
60
 
61
  model = YOLO("platform_yolov11n.pt")
62
- results = model.predict("image.jpg", conf=0.5)
63
  ```
64
 
65
- ### TensorRT Inference
66
 
67
- ```python
68
- from ultralytics import YOLO
69
-
70
- model = YOLO("platform_yolov11n.engine")
71
- results = model.predict("image.jpg")
72
- ```
73
-
74
- ## Training Details
75
-
76
- - **Architecture**: YOLOv11n (nano)
77
- - **Input Size**: 640x640
78
  - **Framework**: Ultralytics
79
 
80
  ## References
81
 
82
- - [IMAV 2025 Official Website](https://femexrobotica.org/imav2025/)
83
- - [IMAV 2025 Rulebook](https://femexrobotica.org/imav2025/index.php/rulebook-imav-2025/)
84
- - [mirela-sdk Detection Module](https://github.com/blackbeedrones/mirela-sdk)
85
-
 
6
  - yolov11
7
  - ultralytics
8
  - drone
9
+ - uav
10
  - imav
11
  - robotics
12
+ - autonomous-landing
13
+ - helipad-detection
14
  datasets:
15
  - custom
16
  pipeline_tag: object-detection
 
18
 
19
  # IMAV 2025 Platform Detection - YOLOv11n
20
 
21
+ Platform detection model for **IMAV 2025 Indoor Competition - Mission 4**.
22
 
23
+ ## Competition Context
24
 
25
+ The [16th International Micro Air Vehicle Conference and Competition (IMAV 2025)](https://femexrobotica.org/imav2025/) took place in San Andrés Cholula, Puebla, Mexico. The competition theme was **"Search and Rescue"**, inspired by Mexico's seismic activity and the need for micro air vehicles in disaster response scenarios.
26
 
27
+ ## Target Object
28
 
29
+ ![Landing Platform](platform.png)
30
+
31
+ **Platform Specifications:**
32
+ - Board: 1m × 1m square
33
+ - Outer circle: Ø 0.85m (black stroke)
34
+ - Inner circle: Ø 0.8m
35
+ - H marking: 0.6m height, 0.35m width, 0.075m stroke
36
+
37
+ ## Mission 4: Land on Moving Platform with Smoke
38
+
39
+ The MAV must autonomously land on a moving platform:
40
+
41
+ | Parameter | Value |
42
+ |-----------|-------|
43
+ | Platform size | 1m × 1m |
44
+ | Lateral movement | up to 1m |
45
+ | Max speed | 0.5 m/s |
46
+ | Obstacle | Smoke machine (partial occlusion) |
47
+
48
+ **Scoring:**
49
 
50
+ | Task | Points |
51
+ |------|--------|
52
+ | No landing | 0 |
53
+ | Landing (stationary) | 2 |
54
+ | Landing (moving platform) | +3 |
55
+ | Landing (with smoke) | +3 |
56
 
57
  ## Model Formats
58
 
 
60
  |--------|------|----------|
61
  | PyTorch | `platform_yolov11n.pt` | Training, fine-tuning |
62
  | ONNX | `platform_yolov11n.onnx` | Cross-platform inference |
63
+ | TensorRT | `platform_yolov11n.engine` | Jetson Orin Nano Super |
64
 
65
  ## Usage
66
 
67
+ ### mirela-sdk
68
 
69
  ```python
70
  from mirela_sdk.ai.detection import Detector
71
 
 
72
  detector = Detector("blackbeedrones/imav-2025-platform:platform_yolov11n.pt")
73
  detector.load()
74
 
 
75
  result = detector.detect(image, conf=0.5)
76
  for det in result:
77
+ print(f"Platform: {det.confidence:.2f} at {det.center}")
78
  ```
79
 
80
+ ### Ultralytics
81
 
82
  ```python
83
  from ultralytics import YOLO
84
 
85
  model = YOLO("platform_yolov11n.pt")
86
+ results = model.predict(image, conf=0.5)
87
  ```
88
 
89
+ ## Training
90
 
91
+ - **Architecture**: YOLOv11n
92
+ - **Input**: 640×640
 
 
 
 
 
 
 
 
 
93
  - **Framework**: Ultralytics
94
 
95
  ## References
96
 
97
+ - [IMAV 2025](https://femexrobotica.org/imav2025/)
98
+ - [Rulebook](https://femexrobotica.org/imav2025/index.php/rulebook-imav-2025/)
99
+ - [mirela-sdk](https://github.com/blackbeedrones/mirela-sdk)