|
|
--- |
|
|
title: compaTAI |
|
|
emoji: 🚗💨 |
|
|
colorFrom: indigo |
|
|
colorTo: blue |
|
|
sdk: static |
|
|
pinned: true |
|
|
license: mit |
|
|
--- |
|
|
|
|
|
# compaTAI: Autonomous Driving for Chaotic Urban Environments 🌆🤖 |
|
|
|
|
|
**Revolutionizing self-driving AI with unstructured urban training data** |
|
|
*"If it works in Mexico City, it works anywhere."* |
|
|
|
|
|
## 🚀 Key Features |
|
|
- **Chaos-proven models**: Specialized in unstructured environments (Mexico City, Mumbai, Lagos) |
|
|
- **Hyperlocal datasets**: 1000+ hours of annotated traffic videos from emerging markets |
|
|
- **Simulation suite**: CARLA-based environment replicating Latin American urban chaos |
|
|
- **Edge case library**: Pedestrians, unmarked roads, aggressive drivers, and more |
|
|
|
|
|
## 📂 Dataset Access |
|
|
```python |
|
|
from datasets import load_dataset |
|
|
mexico_city_traffic = load_dataset("compaTAI/CDMX-Chaos-v5") |
