Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -1,3 +1,98 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- autonomous-driving
|
| 5 |
+
- carla
|
| 6 |
+
- yolo
|
| 7 |
+
- lane-detection
|
| 8 |
+
- ufld
|
| 9 |
+
- pytorch
|
| 10 |
+
- computer-vision
|
| 11 |
+
- self-driving-car
|
| 12 |
+
datasets:
|
| 13 |
+
- custom
|
| 14 |
+
language:
|
| 15 |
+
- en
|
| 16 |
+
---
|
| 17 |
+
|
| 18 |
+
# Autonomous Driving CARLA Models
|
| 19 |
+
|
| 20 |
+
Pre-trained models for autonomous driving in the CARLA 0.9.15 simulator.
|
| 21 |
+
|
| 22 |
+
## Models
|
| 23 |
+
|
| 24 |
+
| Model | File | Size | Description |
|
| 25 |
+
|-------|------|------|-------------|
|
| 26 |
+
| YOLO11n | `yolo_carla_best.pt` | ~6MB | Object detection (vehicles, pedestrians, traffic lights, speed signs) |
|
| 27 |
+
| UFLD | `ufld_carla_best.pth` | ~735MB | Lane detection fine-tuned on CARLA |
|
| 28 |
+
| TuSimple | `tusimple_18.pth` | ~245MB | Pre-trained lane detection backbone |
|
| 29 |
+
|
| 30 |
+
## YOLO11n Object Detection
|
| 31 |
+
|
| 32 |
+
Custom-trained on CARLA simulator data.
|
| 33 |
+
|
| 34 |
+
**Classes:**
|
| 35 |
+
- 0: Vehicle
|
| 36 |
+
- 1: Pedestrian
|
| 37 |
+
- 2: Traffic Light
|
| 38 |
+
- 3: Speed Limit Sign
|
| 39 |
+
|
| 40 |
+
**Training:**
|
| 41 |
+
- Epochs: 250
|
| 42 |
+
- Batch Size: 128
|
| 43 |
+
- Optimizer: AdamW
|
| 44 |
+
- Learning Rate: 0.005
|
| 45 |
+
- Image Size: 640×640
|
| 46 |
+
|
| 47 |
+

|
| 48 |
+

|
| 49 |
+
|
| 50 |
+
## UFLD Lane Detection
|
| 51 |
+
|
| 52 |
+
Fine-tuned Ultra Fast Lane Detection model.
|
| 53 |
+
|
| 54 |
+
**Architecture:**
|
| 55 |
+
- Backbone: ResNet-18
|
| 56 |
+
- Grid: 100×56
|
| 57 |
+
- Input: 800×288
|
| 58 |
+
|
| 59 |
+
**Training:**
|
| 60 |
+
- Epochs: 50
|
| 61 |
+
- Batch Size: 32
|
| 62 |
+
- Optimizer: Adam
|
| 63 |
+
- Learning Rate: 1e-4
|
| 64 |
+
- Loss: SoftmaxFocalLoss + ParsingRelationLoss
|
| 65 |
+
|
| 66 |
+

|
| 67 |
+
|
| 68 |
+
## Usage
|
| 69 |
+
|
| 70 |
+
```python
|
| 71 |
+
from huggingface_hub import hf_hub_download
|
| 72 |
+
|
| 73 |
+
# Download YOLO model
|
| 74 |
+
yolo_path = hf_hub_download(
|
| 75 |
+
repo_id="jkdxbns/autonomous-driving-carla",
|
| 76 |
+
filename="yolo_carla_best.pt"
|
| 77 |
+
)
|
| 78 |
+
|
| 79 |
+
# Download UFLD model
|
| 80 |
+
ufld_path = hf_hub_download(
|
| 81 |
+
repo_id="jkdxbns/autonomous-driving-carla",
|
| 82 |
+
filename="ufld_carla_best.pth"
|
| 83 |
+
)
|
| 84 |
+
```
|
| 85 |
+
|
| 86 |
+
## GitHub Repository
|
| 87 |
+
|
| 88 |
+
Full source code: [github.com/jkdxbns/autonomous-driving-carla](https://github.com/jkdxbns/autonomous-driving-carla)
|
| 89 |
+
|
| 90 |
+
## Author
|
| 91 |
+
|
| 92 |
+
**Justin Mascarenhas**
|
| 93 |
+
CMPE 789 - Robot Perception
|
| 94 |
+
Rochester Institute of Technology
|
| 95 |
+
|
| 96 |
+
## License
|
| 97 |
+
|
| 98 |
+
MIT License
|