Trained 3D Models for CNN based Object Recognition

被引：15

作者：

Sarkar, Kripasindhu ^{[1
,2
]}

Varanasi, Kiran ^{[1
]}

Stricker, Didier ^{[1
,2
]}

机构：

[1] German Res Ctr Artificial Intelligence DFKI, Kaiserslautern, Germany

[2] Tech Univ Kaiserslautern, Kaiserslautern, Germany

来源：

PROCEEDINGS OF THE 12TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS (VISIGRAPP 2017), VOL 5 | 2017年

关键词：

Object Recognition; Fine-tuning CNNs; Domain Fusion; Training on 3D Data; Graphics Assisted CNN;

D O I：

10.5220/0006272901300137

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

We present a method for 3D object recognition in 2D images which uses 3D models as the only source of the training data. Our method is particularly useful when a 3D CAD object or a scan needs to be identified in a catalogue form a given query image; where we significantly cut down the overhead of manual labeling. We take virtual snapshots of the available 3D models by a computer graphics pipeline and fine-tune existing pretrained CNN models for our object categories. Experiments show that our method performs better than the existing local-feature based recognition system in terms of recognition recall.

引用

页码：130 / 137

页数：8

共 50 条

[1] Shape-based Object Recognition in Videos Using 3D Synthetic Object Models
Toshev, Alexander
Makadia, Ameesh
Daniilidis, Kostas
[J]. CVPR: 2009 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, VOLS 1-4, 2009, : 288 - +
[2] Object recognition via MINACE filter trained on synthetic 3D model
Shaulskiy, Dmitry V.
Konstantinov, Maxim V.
Starikov, Rostislav S.
[J]. OPTICS AND PHOTONICS FOR INFORMATION PROCESSING IX, 2015, 9598
[3] Semi-supervised 3D object recognition through CNN labeling
Carlos Rangel, Jose
Martinez-Gomez, Jesus
Romero-Gonzalez, Cristina
Garcia-Varea, Ismael
Cazorla, Miguel
[J]. APPLIED SOFT COMPUTING, 2018, 65 : 603 - 613
[4] Flexible object models for category-level 3D object recognition
Kushal, Akash
Schmid, Cordelia
Ponce, Jean
[J]. 2007 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, VOLS 1-8, 2007, : 1370 - +
[5] LonchaNet: A Sliced-based CNN Architecture for Real-time 3D Object Recognition
Gomez-Donoso, F.
Garcia-Garcia, A.
Garcia-Rodriguez, J.
Orts-Escolano, S.
Cazorla, M.
[J]. 2017 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2017, : 412 - 418
[6] ZigzagNet: Efficient Deep Learning for Real Object Recognition Based on 3D Models
Wang, Yida
Cui, Can
Zhou, Xiuzhuang
Deng, Weihong
[J]. COMPUTER VISION - ACCV 2016, PT IV, 2017, 10114 : 456 - 471
[7] Generative Model With Coordinate Metric Learning for Object Recognition Based on 3D Models
Wang, Yida
Deng, Weihong
[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING, 2018, 27 (12) : 5813 - 5826
[8] VIEW-BASED MODELS OF 3D OBJECT RECOGNITION - INVARIANCE TO IMAGING TRANSFORMATIONS
VETTER, T
HURLBERT, A
POGGIO, T
[J]. CEREBRAL CORTEX, 1995, 5 (03) : 261 - 269
[9] Region Based on Object Recognition in 3D Scenes
Xu, Lei
Zhou, Yue
Li, Qingshan
[J]. INTELLIGENT SCIENCE AND INTELLIGENT DATA ENGINEERING, ISCIDE 2011, 2012, 7202 : 160 - 166
[10] 3D object recognition based on local descriptors
Jakab, Marek
Benesova, Wanda
Racev, Marek
[J]. INTELLIGENT ROBOTS AND COMPUTER VISION XXXII: ALGORITHMS AND TECHNIQUES, 2015, 9406

← 1 2 3 4 5 →