<b>InterHub: A Naturalistic Trajectory Dataset with Dense Interaction for Autonomous Driving</b>

We provide a dense interaction dataset, InterHub, derived from extensive naturalistic driving records to address the scarcity of real-world datasets capturing rich interaction events.The dataset provided on this page include:<b>A</b><b> CSV file </b>(Interactive_Segments_Index.csv) containing the <b>indexed list</b> of the extracted interaction events<b>. </b>In addition to indexing and tracing information about interaction scenarios, we also provide <b>some interesting labels</b> to facilitate more targeted retrieval and utilization of interaction scenarios.(For detailed information, please refer to https://github.com/zxc-tju/InterHub.)<b>Relevant unified data cache files </b>(InterHub_cache_files.zip that includes cache files of lyft_train_full, nuplan_train)<b>.</b>The Python codes used to process and analyze the dataset can be found at https://github.com/zxc-tju/InterHub. The tools for navigating InterHub involve the following three parts:0_data_unify.py converts various data resources into a unified format for seamless interaction event extraction.1_interaction_extract.py extracts interactive segments from unified driving records.2_case_visualize.py showcases interaction scenarios in InterHub.You can refer to the data structure of cache files presented in dataset.md, and after extracting the InterHub_cache_files.zip file, put it in the corresponding folder.<br><b>Statement: </b> <b>All third-party data redistributions included in the interhub_cache_files.zip repository are carried out in full compliance with the original licensing terms of the respective source datasets, as required by their mandatory licensing conditions. This portion of the data remains subject to its original licenses, and users of the data are required to comply with these original licensing terms in any subsequent use or redistribution.</b>