# ICCV2025 BinEgo‑360 Challenge - TAL — Data

来源：https://www.kaggle.com/competitions/bin-ego-360-challenge-tal/data

采集时间：2026-10-18T13:37:52Z

The challenge uses our public dataset 360+x for training/validation, and a held-out test set for the evaluation. The full dataset can be fetched: [https://huggingface.co/datasets/quchenyuan/360x_dataset_HR](https://huggingface.co/datasets/quchenyuan/360x_dataset_HR)


![](assets/36312dec71af8e3a2dc34c2d.gif)

Dataset Overview:
360+x captures real-world environments from four synchronised viewpoints (360° panoramic, binocular egocentric, monocular egocentric, third-person front) together with 6-channel spatial audio, GPS, weather and text metadata. The dataset is designed to study how an embodied agent could perceive, localise actions and recognise scenes when immersed in a space, while also observing itself from an external vantage.