Napredna pretraga

Pregled bibliografske jedinice broj: 972312

An Online Multi-Face Tracker for Unconstrained Videos


Marčetić, Darijan; Ribarić, Slobodan;
An Online Multi-Face Tracker for Unconstrained Videos // Proceedings SITIS 2018
New York, SAD: IEEE Computer Society Conference Publishing Services, 2018. str. 159-165 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
An Online Multi-Face Tracker for Unconstrained Videos

Autori
Marčetić, Darijan ; Ribarić, Slobodan ;

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings SITIS 2018 / - New York, SAD : IEEE Computer Society Conference Publishing Services, 2018, 159-165

ISBN
978-1-5386-9385-8

Skup
2018 14th International Conference on Signal-Image Technology & Internet-Based Systems

Mjesto i datum
Las Palmas de Gran Canaria, Spain, 26-29.11. 2018

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Multi-face tracking, Tracking by detection, Face identity label assignment

Sažetak
This paper presents a system for online multi-face tracking in unconstrained videos. Different shooting angles, strong illumination changes, abrupt motion and face pose changes are the main characteristics of these videos. The proposed online multi-face tracking system combines deep convolutional neural network face detection, multiple instances of a tracker based on discriminative scale and space correlation filters, shot change detection, tracking failure detection, tracklet generation, and ResNet-based face identity label assignment. The system is tested on a dataset of YouTube music videos which is characterised by video sequences with great visual differences caused by face appearance variations (changes in pose, size, makeup, and illumination), and/or rapid camera motion. The results of the experiment expressed by MOTA, MOTP and IDS metrics are given and compared with state-of-the-art multi-target trackers.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo



POVEZANOST RADA


Projekt / tema
6733 HRZZ
7619 HRZZ

Ustanove
Fakultet elektrotehnike i računarstva, Zagreb

Profili:

Avatar Url Slobodan Ribarić (autor)

Avatar Url Darijan Marčetić (autor)

Citiraj ovu publikaciju

Marčetić, Darijan; Ribarić, Slobodan;
An Online Multi-Face Tracker for Unconstrained Videos // Proceedings SITIS 2018
New York, SAD: IEEE Computer Society Conference Publishing Services, 2018. str. 159-165 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
Marčetić, D., Ribarić, S. & (2018) An Online Multi-Face Tracker for Unconstrained Videos. U: Proceedings SITIS 2018.
@article{article, year = {2018}, pages = {159-165}, keywords = {Multi-face tracking, Tracking by detection, Face identity label assignment}, isbn = {978-1-5386-9385-8}, title = {An Online Multi-Face Tracker for Unconstrained Videos}, keyword = {Multi-face tracking, Tracking by detection, Face identity label assignment}, publisher = {IEEE Computer Society Conference Publishing Services}, publisherplace = {Las Palmas de Gran Canaria, Spain} }

Časopis indeksira:


  • Web of Science Core Collection (WoSCC)
    • Conference Proceedings Citation Index - Science (CPCI-S)
  • Scopus