Sequential Image-based Attention Network for Inferring Interaction Force without Haptic sensor

DC Field Value Language
dc.description학위논문(석사)--아주대학교 일반대학원 :컴퓨터공학과,2019. 2-
dc.description.tableofcontentsⅡ. Introduction 1 Ⅲ. Related Work 6 Ⅳ. Proposed Method 8 A. Baseline: CNN-based RNN method for sequential image description 8 1. Visual Feature Extraction 8 2. LSTM Sequential Model 9 B. Weighted Average Pooling (WAP) 9 C. Sequential Spatial Attention Module(SSAM) 11 D. Sequential Channel Attention Module(SCAM) 13 E. Ensemble Module 14 Ⅴ. Dataset and Implementation 16 A. Experimental Setup and Database 16 B. Implementation Detail 18 Ⅵ. Experimental Results and Discussion 20 A. Experimental Results on Proposed Sequential Attention Module 20 B. Experimental Result on different network architecture 21 C. Comparative Evaluation with well-known method 22 D. Performance analysis according to force intensity changes 23 E. Performance analysis on Various Material 24 Conclusion 26 Reference 29-
dc.publisherThe Graduate School, Ajou University-
dc.rights아주대학교 논문은 저작권에 의해 보호받습니다.-
dc.titleSequential Image-based Attention Network for Inferring Interaction Force without Haptic sensor-
dc.contributor.affiliation아주대학교 일반대학원-
dc.contributor.department일반대학원 컴퓨터공학과- 2-
dc.description.alternativeAbstractHumans can infer approximate interaction force between objects from only vision information because we already have learned it through experiences. Based on this idea, we propose a recurrent convolutional-neural-network-based method using sequential images for inferring interaction force without using a haptic sensor. For training and validating deep learning methods, we collected a large number of (e.g. 359,413) images and corresponding interaction forces through an electronic motor-based device. To concentrate on changing shapes of a target object by the external force in images, we propose a sequential image-based attention module, which learns a salient model from temporal dynamics. The proposed sequential image-based attention module consists of a sequential spatial attention module and a sequential channel attention module, which are extended to exploit multiple sequential images. For gaining better accuracy, we also created a weighted average pooling layer for both spatial and channel attention modules. The extensive experimental results verified that the proposed method successfully infers interaction forces under the various conditions, such as different target materials, illumination changes, and external force directions.-
Appears in Collections:
Graduate School of Ajou University > Department of Computer Engineering > 3. Theses(Master)
Files in This Item:
There are no files associated with this item.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.