Sign inGet started
Dataset · Hugging Face

mrhwl123/MOT17

MOT17 MOT17 是 MOTChallenge 多目标跟踪基准的 2017 年版,是行人跟踪领域使用最广泛的评测集之一,聚焦拥挤街景下的遮挡与身份保持问题。 数据构成 14 个视频序列(7 个训练 + 7 个测试),每个序列以 3 种公开检测器结果提供(DPM / FRCNN / SDP),因此实际共有 21 个可训练序列。 视频帧率通常为 30 FPS(部分序列剪辑自更高帧率源)。 图像来自固定与移动相机视角的公开/半公开街景场景。 目录结构 {MOT17 ROOT} ├── train/ │ └── MOT17-02-FRCNN/ │ ├── img1/ # 图像序列 │ ├── det/det.txt # 公开检测结果 │ └── gt/gt.txt # 标注(仅训练集提供) └── test/ └── MOT17-01-FRCNN/ ├── img1/… See the full description on the dataset page: https://huggingface.co/datasets/mrhwl123/MOT17.

Author / organizationmrhwl123
License (as declared)cc-by-nc-sa-3.0
Downloads400 (as reported by Hugging Face)
Likes0
Created25 August 2026
Last updated at source15 September 2026
Embed this entry
<iframe src="https://silytics.si/embed/item/a98a7e62-1330-46d4-bf69-48c938cb098d" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 3 days

400 now (+108 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

More from mrhwl123

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.