BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Fast Tagging of Pollinator Field Videos with Convolutional Tsetlin
  Machines - Sachin Mathew\, University of Cambridge
DTSTART:20230922T120500Z
DTEND:20230922T130000Z
UID:TALK204421@talks.cam.ac.uk
CONTACT:114742
DESCRIPTION:Entomologists devote a large portion of their time manually ta
 gging video data from camera traps in order to conduct their research. Thi
 s is an enormous time\, labor\, and resource sink. Automation would greatl
 y decrease the amount of work required to complete this task and would giv
 e these researchers the freedom to allocate their resources elsewhere. Des
 pite the difficulty of this task due to the comparable scale of the insect
 s and visual noise\, the structure of these static camera videos lends its
 elf to be interpreted by sufficiently robust machine learning models.\nThi
 s work aims to address the task of tagging location specific events within
  insect camera traps --- such as pollination events --- at real-time or cl
 ose to real-time speeds by implementing a pipeline of fast video regulariz
 ation\, background subtraction\, and machine learning inference using the 
 highly parallelizable and embeddable Convolutional Tsetlin Machine (CTM) a
 rchitecture. This work presents a pipeline of fast regularization and back
 ground subtraction models\, compared by the metrics of event detection rat
 e\, pollination image detection rate as well as pipeline iteration speed. 
 Through this exploration a subset of operations were found such that indiv
 idual pollination events were tagged with relative accuracy at fast rates 
 with outputs easily interpretable by CTMs for an even higher detection rat
 e at very high possible speeds in embedded systems.\n\nSachin Mathew is a 
 Research Associate in the Department of Computer Science at the University
  of Cambridge. Their research focuses on developing non-invasive methods f
 or gathering insect/small animal data for biodiversity and conservation ta
 sks using computer vision.
LOCATION:FW 11\, William Gates Building. Zoom link: https://cl-cam-ac-uk.z
 oom.us/j/4361570789?pwd=Nkl2T3ZLaTZwRm05bzRTOUUxY3Q4QT09&amp\;from=addon 
END:VEVENT
END:VCALENDAR