CHIRP Dataset: Combining beHaviors, Individual Re-identification and Postures

DOI

CHIRP is a computer vision dataset on Siberian Jays (Perisoreus infaustus) living in the Swedish Laplands. With the goal of long-tem individual level behavioural monitoring, we provide annotations for a diverse set of computer vision tasks, including video re-identification, action recognition, 2D keypoint estimation, instance segmentation and object detection. We also provide an application specific benchmark to benchmark algorithms directly with the final biological measurement. Code and info for accessing the dataset: https://github.com/alexhang212/CHIRP_Dataset

Identifier
DOI https://doi.org/10.17617/3.GVO4LG
Metadata Access https://edmond.mpg.de/api/datasets/export?exporter=dataverse_json&persistentId=doi:10.17617/3.GVO4LG
Provenance
Creator Chan, Hoi Hang; Singhal, Neha; Kocahan, Onur; Meltzer, Andrea; Lubrano, Saverio; Warrington, Miyako; Griesser, Michael; Kano, Fumihiro; naik, Hemal
Publisher Edmond
Publication Year 2026
Funding Reference Germany’s Excellence Strategy—EXC 2117—422037984; 15990824; Heisenberg Grant no. GR 4650/21; FP589/20; Deutsche Forschungsgemeinschaft
OpenAccess true
Contact achan(at)ab.mpg.de; hnaik(at)ab.mpg.de
Representation
Language English
Resource Type Dataset
Version 1
Discipline Other