TY - GEN
T1 - Aggregated Deep Convolutional Neural Networks for Multi-View 3D Object Retrieval
AU - Alzu'bi, Ahmad
AU - Abuarqoub, Abdelrahman
AU - Al-Hmouz, Ahmed
N1 - Publisher Copyright:
© 2019 IEEE.
PY - 2019/10/30
Y1 - 2019/10/30
N2 - Extracting and aggregating discriminative image features is a key challenge for 3D multi-view object recognition and retrieval tasks. In this paper, we propose aggregated deep CNNs (ADCNN) model to address the limitations associated with projecting 3D models into multiple 2D images and their resulting high-dimensional representations. A deep learning network is developed to aggregate compact features of 3D objects using the activation kernels of convolutional layers directly. Two instances of the same CNN features extractor share the learning weights while they represent different object characteristics. Systematic experiments conducted on the benchmark dataset ModelNet40 demonstrate the efficacy of the proposed method in 3D object retrieval and a m AP accuracy of 91.1% is achieved, which shows its performance superiority over related state-of-the-art methods.
AB - Extracting and aggregating discriminative image features is a key challenge for 3D multi-view object recognition and retrieval tasks. In this paper, we propose aggregated deep CNNs (ADCNN) model to address the limitations associated with projecting 3D models into multiple 2D images and their resulting high-dimensional representations. A deep learning network is developed to aggregate compact features of 3D objects using the activation kernels of convolutional layers directly. Two instances of the same CNN features extractor share the learning weights while they represent different object characteristics. Systematic experiments conducted on the benchmark dataset ModelNet40 demonstrate the efficacy of the proposed method in 3D object retrieval and a m AP accuracy of 91.1% is achieved, which shows its performance superiority over related state-of-the-art methods.
KW - 3D object retrieval
KW - compact pooling
KW - convolutional neural networks
KW - deep learning
UR - http://www.scopus.com/inward/record.url?scp=85079099649&partnerID=8YFLogxK
U2 - 10.1109/ICUMT48472.2019.8970827
DO - 10.1109/ICUMT48472.2019.8970827
M3 - Conference contribution
AN - SCOPUS:85079099649
T3 - International Congress on Ultra Modern Telecommunications and Control Systems and Workshops
BT - 11th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops, ICUMT 2019
PB - IEEE Computer Society
T2 - 11th International Congress on Ultra Modern Telecommunications and Control Systems and Workshops, ICUMT 2019
Y2 - 28 October 2019 through 30 October 2019
ER -