遇见数据集

Innate heuristics and fast learning support escape route selection in mice

收藏
Zenodo2022-04-25 更新2026-05-25 收录
数据链接:
官方服务:

资源简介:

### Dataset structure Each experimental session (from one mouse on one arena) was composed of one or more recording. Each recording<br> was saved as a separate video and as such the tracking data and stimuli metadata are saved separately for each<br> recording. The tracking data for each recording is saved in a file `DATE_SESSION_RECORDING_tracking.h5` with DLC pose position for<br> four body parts: nose, neck, body and base of the tail.<br> The video data was acquired at 30 or 40 FPS, 30FPS recordings were upsampled so that all tracking data<br> saved here is at 40 FPS.<br> If `_recording` is missing from the filename, the session had only one recording. The `DATE_SESSION_RECORDING_stimuli.csv` files store stimuli metadata for each recording.<br> The columns `overview_frame` and `overview_frame_end` denote stimulus onset and offset. The `session_metadata.csv` stores metadata about all sessions and recordings in the data set.<br> Metadata include mouse name, arena type (1-5) and if the mouse was naive or not (1/0). <br> For the analyses in the manuscript, stimulus times were used to define trials and the tracking data in the few seconds<br> after the stimulus onbset was analzed (from stimulus onset to mouse entering the shelter). For convenience we provide<br> `trials.h5` which stores all tracking data and metadata for all trials and `trials_metadata.csv` that stores the metadata only.

提供机构:
Zenodo
创建时间:
2022-04-25
二维码
社区交流群
二维码
科研交流群
商业服务