Original title:
Detekce obličeje v nekvalitních videozáznamech
Translated title:
Face Detection in Poor Quality Videos
Authors:
Koval, Michal ; Orság, Filip (referee) ; Goldmann, Tomáš (advisor) Document type: Bachelor's theses
Year:
2021
Language:
slo Publisher:
Vysoké učení technické v Brně. Fakulta informačních technologií Abstract:
[slo][eng]
Táto bakalárska práca sa venuje problematike detekcie tvárí v nekvalitných videozáznamoch, pričom sa konkrétne zameriava na prekryté tváre. Popisuje základné princípy algoritmov strojového učenia a ich métody, ktoré sú často využívané v oblasti počítačového videnia. Z nich sú bližšie priblížené konvolučné neurónové siete a ich state of the art modely zamerané detekciu tvárí. V praktickej časti boli vytvorené a natrénované modely na detekciu tvárí inšpirované známym state of the art modelom RetinaFace. Najlepšia varianta z nich dosauje na WIDER Face HARD testovacom sete 85,5% priemernú presnosť a na testovacom sete zameranom na prekryté tváre 90,9%. Súčasťou práce je aj program s grafickým uživateľským rozhraním, ktorý poskytuje nástroje na použitie vytrénovaných modelov na videu a obrázkoch.
This bachelor thesis deals with face detection in low quality videos, while mainly focusing on occluded faces. It describes elementary priciples of machine learning algorithms and their methods, which are often used in the field of computer vision. Out of them are more closely described convolutional neural networks and their state of the art models focused on face detection. Out of those, convolutional neural networks and state of the art models for face detection are more closely described. For the practical part face detection models inspired by state of the art model RetinaFace were implemented and trained. The best performing model achieves 85.5% average precision on WIDER Face HARD testing dataset and 90.9% on dataset focused on occluded faces. Part of this thesis is also a program with graphical user interfaces which provides tools to use developed models on videos and pictures.
Keywords:
closed circuit cameras; convolutional neural networks; face detection; Keras; machine learning; neural networks; occluded faces; poor quality videos; RetinaFace; WIDER Face
Institution: Brno University of Technology
(web)
Document availability information: Fulltext is available in the Brno University of Technology Digital Library. Original record: http://hdl.handle.net/11012/199312