PT  - JOURNAL ARTICLE
AU  - Francisco Romero-Ferrero
AU  - Mattia G. Bergomi
AU  - Robert Hinz
AU  - Francisco J. H. Heras
AU  - Gonzalo G. de Polavieja
TI  - idtracker.ai: Tracking all individuals in large collectives of unmarked animals
AID  - 10.1101/280735
DP  - 2018 Jan 01
TA  - bioRxiv
PG  - 280735
4099  - http://biorxiv.org/content/early/2018/03/14/280735.short
4100  - http://biorxiv.org/content/early/2018/03/14/280735.full
AB  - Our understanding of collective animal behavior is limited by our ability to track each of the individuals. We describe an algorithm and software, idtracker.ai, that extracts from video all trajectories with correct identities at a high accuracy for collectives of up to 100 individuals. It uses two deep networks, one detecting when animals touch or cross and an-other one for animal identification, trained adaptively to conditions and difficulty of the video.