Integrated Inference and Learning of Neural Factors in Structural Support Vector Machines

Autor: Rein Houthooft, Filip De Turck
Jazyk: angličtina
Rok vydání: 2015
Předmět:
FOS: Computer and information sciences
Computer science
Computer Vision and Pattern Recognition (cs.CV)
FEATURES
Computer Science - Computer Vision and Pattern Recognition
Inference
Machine Learning (stat.ML)
02 engineering and technology
computer.software_genre
Machine learning
CLASSIFICATION
Machine Learning (cs.LG)
030507 speech-language pathology & audiology
03 medical and health sciences
Artificial Intelligence
Margin (machine learning)
Statistics - Machine Learning
0202 electrical engineering
electronic engineering
information engineering
Neural and Evolutionary Computing (cs.NE)
Structured prediction
SCENE
Subgradient method
Image segmentation
Artificial neural network
business.industry
Computer Science - Neural and Evolutionary Computing
Neural factors
Support vector machine
Computer Science - Learning
Mathematics and Statistics
MULTICLASS OBJECT RECOGNITION
Signal Processing
Pattern recognition (psychology)
020201 artificial intelligence & image processing
IBCN
Computer Vision and Pattern Recognition
Artificial intelligence
Data mining
MINIMIZATION
Structural support vector machine
0305 other medical science
business
computer
Software
Neural networks
Zdroj: PATTERN RECOGNITION
ISSN: 0031-3203
Popis: Tackling pattern recognition problems in areas such as computer vision, bioinformatics, speech or text recognition is often done best by taking into account task-specific statistical relations between output variables. In structured prediction, this internal structure is used to predict multiple outputs simultaneously, leading to more accurate and coherent predictions. Structural support vector machines (SSVMs) are nonprobabilistic models that optimize a joint input-output function through margin-based learning. Because SSVMs generally disregard the interplay between unary and interaction factors during the training phase, final parameters are suboptimal. Moreover, its factors are often restricted to linear combinations of input features, limiting its generalization power. To improve prediction accuracy, this paper proposes: (i) joint inference and learning by integration of back-propagation and loss-augmented inference in SSVM subgradient descent; (ii) extending SSVM factors to neural networks that form highly nonlinear functions of input features. Image segmentation benchmark results demonstrate improvements over conventional SSVM training methods in terms of accuracy, highlighting the feasibility of end-to-end SSVM training with neural factors. HighlightsA novel structured prediction model is proposed and applied to image segmentation.SSVM factors are modeled by highly nonlinear functions through neural networks.Back-propagation and loss-augmented inference are integrated in subgradient descent.Segmentation benchmark accuracy results show benefits over standard SSVM methods.
Databáze: OpenAIRE
Externí odkaz: