Authors:
Luca Bergamini
1
;
Stefano Pini
1
;
Alessandro Simoni
1
;
Roberto Vezzani
1
;
Simone Calderara
1
;
Rick B. D’Eath
2
and
Robert B. Fisher
3
Affiliations:
1
University of Modena and Reggio Emilia, Italy
;
2
SRUC, Edinburgh, U.K.
;
3
University of Edinburgh, U.K.
Keyword(s):
Pig Detection, Pig Tracking, Behavior Classification, Pig Farming, Long-term Temporal Analysis.
Abstract:
Visual observation of uncontrolled real-world behavior leads to noisy observations, complicated by occlusions, ambiguity, variable motion rates, detection and tracking errors, slow transitions between behaviors, etc. We show in this paper that reliable estimates of long-term trends can be extracted given enough data, even though estimates from individual frames may be noisy. We validate this concept using a new public dataset of approximately 20+ million daytime pig observations over 6 weeks of their main growth stage, and we provide annotations for various tasks including 5 individual behaviors. Our pipeline chains detection, tracking and behavior classification combining deep and shallow computer vision techniques. While individual detections may be noisy, we show that long-term behavior changes can still be extracted reliably, and we validate these results qualitatively on the full dataset. Eventually, starting from raw RGB video data we are able to both tell what pigs main daily
activities are, and how these change through time.
(More)