BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Through a glass darkly - interpreting animal behaviour with the ai
 d of machines - Alex Jordan (Max Planck Institute for Ornithology)
DTSTART:20190205T130000Z
DTEND:20190205T140000Z
UID:TALK117520@talks.cam.ac.uk
CONTACT:Arne Jungwirth
DESCRIPTION:The qualitative and potentially subjective nature of many etho
 logical studies can lead to unresolvable debate over the interpretation of
  animal behaviour. This problem can be exacerbated when the taxonomic dist
 ance between the human observer and the focal species increases. In this t
 alk I will discuss my recent work examining the potential for self-recogni
 tion in fish using the mirror test. In the study\, we show that a fish\, t
 he cleaner wrasse Labroides dimidiatus\, shows behaviour that may reasonab
 ly be interpreted as passing through all phases of the mark test: (i) soci
 al reactions towards the reflection\, (ii) repeated idiosyncratic behaviou
 rs towards the mirror\, and (iii) frequent observation of their reflection
 . When subsequently provided with a coloured tag in a modified mark test\,
  fish attempt to remove the mark by scraping their body in the presence of
  a mirror but show no response towards transparent marks or to coloured ma
 rks in the absence of a mirror. This study has been met with significant r
 esistance from parts of the behaviour community\, who argue that interpret
 ations of fish behaviour cannot be made in the same way as for mammals. I 
 will discuss how our current approaches employing machine vision and artif
 icial neural networks may provide a more objective and quantitative descri
 ption of animal behaviours that open greater avenues to debate and discuss
 ion based on data rather than intuition.
LOCATION:Part II Lecture Theatre\, Department of Zoology \, Downing Street
 \, CB2 3EJ
END:VEVENT
END:VCALENDAR