声明
摘要
ABSTRACT
Contents
List of Figures
List of Tables
List of Algorithms
Chapter 1 Introduction
1.1 Research Background and Significance
1.2 Research Content
1.3 Organization
Chapter 2 Literature Review
2.1 Introduction
2.2 Datasets
2.2.1.The Weizmann Human Action Dataset
2.2.2.The KTH Human Action Dataset
2.2.3.The UCF-101 Action Recognition Dataset
2.2.4.The HMDB-51 Dataset
2.3 Global Features Representation
2.4 Local Features Representation
2.5 Binary Motion Feature Extraction(Dynamic Texture)
2.5.3.Combination of Binary Descriptors with Floating-Point Descriptors
2.6 Deep Learning Architectures
2.6.1. 3D CNN Networks
2.6.2.Two-Stream Networks
2.6.3.Temporal Dynamic Modeling with Temporal Pooling
2.6.4.Temporal Evolution Captured with RNN
2.7 Summary
Chapter 3 Binary Motion Description for Action Recognition in Videos
3.1 Introduction
3.2 The Proximity Patches Pattern
3.3 BPPEM Descriptor
3.3.1.Overview
3.3.2.Computation of BPPEM
3.4 Proximity Patches Similarity Motion Descriptor
3.4.1. Introduction to PPSM
3.4.2.Computation of PPSM
3.5 Experiment Setup
3.5.1.Framework,Hardware and Software Specifications
3.5.2. Evaluation Metrics
3.6 Results and Analysis
3.6.1.Number of Surrounding Patches
3.6.2. SSD vs FND
3.6.3.Temporal Distance Between two Consecutive Frames
3.6.4. BPPEM
3.6.5. eBPPEM
3.6.6.PPSM
3.6.7. ePPSM
3.6.8. BPPEM-PPSM,and eBPPEM-ePPSM Fusions
3.6.9. Comparision with the State-of-the-art
3.7 Summary
Chapter 4 Spatial Binary Descriptors for Human Action Recognition
4.1 Introduction
4.2 FREAK,BinBoost,LATCH
4.3 Action Recognition with FREAK,BinBoost,LATCH Appearance Descriptors
4.4 Binary Spatio-Temporal Descriptors with FREAK 8,BinBoost 16 and LATCH 8 as Appearance Descriptors
4.4.1.Analysis
4.5 Summary
Chapter 5 3D Spatio-Temporal Binary CNNs
5.1 Introduction
5.2 Related Works
5.2.1. 3D Convolutional Networks
5.3 Proposed Model:3D Spatio-Temporal Binary Convolutional Network
5.3.1. Binarized ConvNets
5.4 3D Spatio-Temporal Binary CNNs(3D ST-BCNN)
5.4.1.Basic Components of the 3D Spatio-Temporal Binary CNNs
5.4.2.Binary Operations
5.4.3.Proposed Framework
5.5 Experimental Results and Analysis
5.5.1. Evaluation with Train and Validation Sets
5.5.2 Evaluation with Train,Validation and Test Sets
5.6 Summary
Chapter 6 Conclusion and Future Works
6.1 Summary
6.2 Future Works
Bibliography
Acknowledgements
Publications