BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Europe/Skopje
BEGIN:DAYLIGHT
DTSTART:20250330T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20241027T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20241209T024009Z
UID:DB9225E8-B94F-4E60-9167-50F797A05E76
DTSTART;TZID=Europe/Skopje:20241127T130000
DTEND;TZID=Europe/Skopje:20241127T140000
DESCRIPTION:Improving the explainability of Computer Vision models based on
  Deep Learning has recently become a compelling problem\, providing reliab
 le predictions for the end user and enabling more fine-grained classificat
 ions. Recently\, Concept Bottleneck models have been proposed for image cl
 assification\, dividing the problem into two stages and thereby defining a
  hierarchy of concepts. In this lecture\, we will present our contribution
  to the topic. We developed a conceptual bottleneck model for image segmen
 tation\, objects fine classification\, and object tracking\, which was tra
 ined and tested on a dataset consisting of images of refrigerators filled 
 with various objects. The proposed model makes full use of the hierarchy i
 n concepts\, exploiting the relationships between different categories at 
 the same hierarchical level and relying on a new method for handling multi
 -label classifications. We will discuss the findings of our research\, esp
 ecially the explainability of the proposed approach\, and elaborate on the
  possibility of future improvements.\n\nCo-sponsored by: FEEIT\, FCSE\, Ex
 plainable AI Project\, EuroCC Project\, ICT-ACT\n\nSpeaker(s): Vesna Dimit
 rievska\, PhD\, \n\nRoom: AMF-FCSE\, University Ss. Cyril and Methodius\, 
 Faculty of Computer Science and Engineering\, Rudzer Boshkovikj 16\, Skopj
 e\, Macedonia\, Macedonia\, 1000
LOCATION:Room: AMF-FCSE\, University Ss. Cyril and Methodius\, Faculty of C
 omputer Science and Engineering\, Rudzer Boshkovikj 16\, Skopje\, Macedoni
 a\, Macedonia\, 1000
ORGANIZER:vojleska@feit.ukim.edu.mk
SEQUENCE:51
SUMMARY:Explainable Machine Learning via Concept Bottleneck
URL;VALUE=URI:https://events.vtools.ieee.org/m/447850
X-ALT-DESC:Description: <br /><div>\n<p><span style="font-size: 11.0pt\; fo
 nt-family: 'Arial'\,sans-serif\; mso-fareast-font-family: Calibri\; mso-fa
 reast-theme-font: minor-latin\; color: #242424\; mso-ansi-language: EN-US\
 ; mso-fareast-language: EN-US\; mso-bidi-language: AR-SA\;">Improving the 
 explainability of Computer Vision models based on Deep Learning has recent
 ly become a compelling problem\, providing reliable predictions for the en
 d user and enabling more fine-grained classifications. Recently\, Concept 
 Bottleneck models have been proposed for image classification\, dividing t
 he problem into two stages and thereby defining a hierarchy of concepts. I
 n this lecture\, we will present our contribution to the topic. We develop
 ed a conceptual bottleneck model for image segmentation\, objects fine cla
 ssification\, and object tracking\, which was trained and tested on a data
 set consisting of images of refrigerators filled with various objects. The
  proposed model makes full use of the hierarchy in concepts\, exploiting t
 he relationships between different categories at the same hierarchical lev
 el and relying on a new method for handling multi-label classifications. W
 e will discuss the findings of our research\,&nbsp\;especially the&nbsp\;e
 xplainability&nbsp\;of the proposed approach\,&nbsp\;and elaborate on the 
 possibility of future improvements.</span></p>\n</div>\n<div>&nbsp\;</div>
END:VEVENT
END:VCALENDAR