Voxel51/NYC_Smells
收藏Hugging Face2025-12-18 更新2025-12-20 收录
下载链接：
https://hf-mirror.com/datasets/Voxel51/NYC_Smells
下载链接
链接失效反馈官方服务：
资源简介：
New York Smells是一个大规模多模态数据集，将视觉图像与电子鼻（e-nose）嗅觉信号配对，这些信号是在纽约市的各种环境中捕获的。该数据集支持嗅觉和视觉之间的跨模态学习，填补了机器感知研究中嗅觉感知的空白。数据集包含7,000个嗅觉-图像对，来自3,500个不同的对象，覆盖60个记录会话。数据集由哥伦比亚大学、康奈尔大学和Osmo Labs的研究团队收集。
New York Smells is a large-scale multimodal dataset pairing visual imagery with electronic nose (e-nose) olfactory signals captured in the wild throughout New York City. The dataset enables cross-modal learning between smell and sight, addressing a critical gap in machine perception research where olfaction remains largely inaccessible to machines compared to vision, sound, and touch. The dataset contains 7,000 smell-image pairs from 3,500 distinct objects across 60 recording sessions. The dataset was curated by a research team from Columbia University, Cornell University, and Osmo Labs.
提供机构：
Voxel51
