Mixed Mode Cascaded Classification Models
نویسنده
چکیده
We present here Mixed Mode Cascaded Classification Models, an algorithmic framework that seeks to effectively solve a wide range of machine learning tasks in a “plug and play” manner. It does this by sharing predictions between machine learning tasks, thus giving each task additional high-level information that can be used to solve its specific problem. We consider here a specific implementation of this framework in which we combine the machine vision tasks of scene categorization and depth estimation. In addition, we consider the use of a Poisson distribution for depth estimation. Experimental results are provided.
منابع مشابه
Mixed-Mode Stress Intensity Factors for Surface Cracks in Functionally Graded Materials Using Enriched Finite Elements
Three-dimensional enriched finite elements are used to compute mixed-mode stress intensity factors (SIFs) for three-dimensional cracks in elastic functionally graded materials (FGMs) that are subject to general mixed-mode loading. The method, which advantageously does not require special mesh configuration/modifications and post-processing of finite element results, is an enhancement of previou...
متن کاملConduction and Dead-Time Voltage Drops Estimation of Asymmetric Cascaded H-Bridge Converters Utilizing Level-Shifted PWM Scheme
Linear AC power supplies can be replaced by their nonlinear switching counterparts due to the lower voltage drops and higher efficiency and power density of switching-mode inverters. Multilevel cascaded H-bridge (CHB) converters are the preferred inverter structure because of modular configuration, control, and protection. The output voltage quality in CHB converters depends on the number of ou...
متن کاملCascaded model adaptation for dialog act segmentation and tagging
There are many speech and language processing problems which require cascaded classification tasks. While model adaptation has been shown to be useful in isolated speech and language processing tasks, it is not clear what constitutes system adaptation for such complex systems. This paper studies the following questions: In cases where a sequence of classification tasks is employed, how importan...
متن کاملLearning cascaded latent variable models for biomedical text classification
In this paper, we develop a weakly supervised version of logistic regression to help to improve biomedical text classification performance when there is limited annotated data. We learn cascaded latent variable models for the classification tasks. First, with a large number of unlabelled but limited amount of labelled biomedical text, we will bootstrap and semi-automate the annotation task with...
متن کاملOptical-frequency balanced mixer.
Optical signal processing devices based on quasi-phase-matched three-wave mixing and cascaded three-wave mixing in guided-wave geometries have been demonstrated to operate efficiently at practical pump-power levels. We describe operation of such devices in a balanced mode that allows mixing without wavelength offset and separation of mixed output from pump and signal input without wavelength-se...
متن کامل