Naslov
On the Comparison of Classic and Deep Keypoint Detector and Descriptor Methods

Autori
Bojanić, David ; Bartol, Kristijan, Pribanić, Tomislav ; Petković, Tomislav ; Diez, Yago ; Salvi, Joaquim

Vrsta, podvrsta i kategorija rada
Sažeci sa skupova, sažetak, znanstveni

Izvornik
Book of Abstracts 8th Croatian Computer Vision Workshop (CCVW2019) / Lončarić, Sven ; Kalafatić, Zoran - Zagreb : Fakultet elektrotehnike i računarstva Sveučilišta u Zagrebu, 2019, 19-20

Skup
8th Croatian Computer Vision Workshop (CCVW2019)

Mjesto i datum
Zagreb, Hrvatska, 23.09.2019

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Nije recenziran

Ključne riječi
keypoint detector ; keypoint descriptor ; SIFT, SURF, ORB

Sažetak
The purpose of this study is to give a performance comparison between several classic hand-crafted and deep keypoint detector and descriptor methods. In particular, we consider the following classical algorithms: SIFT, SURF, ORB, FAST, BRISK, MSER, HARRIS, KAZE, AKAZE, AGAST, GFTT, FREAK, BRIEF and RootSIFT, where a subset of all combinations is paired into detectordescriptor pipelines. Additionally, we analyze the performance of two recent and perspective deep detector-descriptor models, LF-Net and SuperPoint. Our benchmark relies on the HPSequences dataset that provides real and diverse images under various geometric and illumination changes. We analyze the performance on three evaluation tasks: keypoint verification, image matching and keypoint retrieval. The results show that certain classic and deep approaches are still comparable, with some classic detector-descriptor combinations overperforming pretrained deep models. In terms of the execution times of tested implementations, SuperPoint model is the fastest, followed by ORB.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Računarstvo

POVEZANOST RADA