OmniTraffic-benchmark-v1

🚗 OmniTraffic: A Large-scale Multi-view Spatiotemporal Benchmark for Traffic Understanding 📌 Benchmark Summary OmniTraffic is a comprehensive evaluation benchmark designed to test the multi-view spatiotemporal reasoning and Bird’s-Eye View (BEV) perception capabilities of multimodal large language models (MLLMs) and autonomous driving systems. While the complete OmniTraffic dataset ecosystem contains an underlying pool of over 8 million generated VQA samples, this… See the full description on the dataset page: Benchmark.

Source: Hugging Face Hub (CROHuang/OmniTraffic_Benchmark). Metadata imported from the dataset’s Hub tags.