Conditional Random Rields (CRF) have been widely applied in image segmentations. While most studies rely on handcrafted features, we here propose to exploit a pre-trained large convolutional neural network (CNN) to generate deep features for CRF learning. The deep CNN is trained on the ImageNet dataset and transferred to image segmentations here for constructing potentials of superpixels. Then ...

Abstract The twin support vector machine improves the classification performance of by solving two small quadratic programming problems. However, this method has following defects: (1) For and some its variants, constructed models use a hinge loss function, which is sensitive to noise unstable in resampling. (2) need be converted from original space dual space, their time complexity high. To fu...

‎In this paper‎, ‎we introduce the structure of a groupoid associated to a vector field‎ ‎on a smooth manifold‎. ‎We show that in the case of the $1$-dimensional manifolds‎, ‎our‎ ‎groupoid has a‎ ‎smooth structure such that makes it into a Lie groupoid‎. ‎Using this approach‎, ‎we associated to‎ ‎every vector field an equivalence‎ ‎relation on the Lie algebra of all vector fields on the smooth...

The Spanish National Cancer Research Center (CNIO) and University of Navarra organized a challenge on recognizing chemical compounds and drugs (chemical entities) in biomedical literature, which includes two individual subtasks: 1) chemical entity mention recognition (CEM); and 2) chemical document indexing (CDI). The challenge organizers manually annotated chemical entities in 10000 abstracts ...

Motivation: A popular approach for predicting RNA secondary structure is the thermodynamic nearest neighbor model that finds a thermodynamically most stable secondary structure with the minimum free energy (MFE). For further improvement, an alternative approach that is based on machine learning techniques has been developed. The machine learning based approach can employ a fine-grained model th...

Based on the fractal dimension, the tri-plots can classify two large and not equal sizes of the time series datasets. The tri-plots measure three function values which include two self-plots and one cross-plot. The self-plot affords the character of one individual dataset. The cross-plot describes the relation between two datasets. Originally, the tri-plots just can get the relation in two data...

The standard approach to recognizing text in images consists in first classifying local image regions into candidate characters and then combining them with high-level word models such as conditional random fields (CRF). This paper explores a new paradigm that departs from this bottom-up view. In our approach, every label from a lexicon is embedded to an Euclidean vector space. We refer to this...

Recent attempts to assess the performance of SSVM algorithms for unconstrained minimization problems differ in their evaluations from earlier assessments. Nevertheless, the new experiments confirm earlier observations that, on certain types of problems, the SSVM algorithms are far superior to other variable metric methods. This paper presents a critical review of these recent assessments and di...

background: this paper proposes a new emotional stress assessment system using multi-modal bio-signals. electroencephalogram (eeg) is the reflection of brain activity and is widely used in clinical diagnosis and biomedical research. methods: we design an efficient acquisition protocol to acquire the eeg signals in five channels (fp1, fp2, t3, t4 and pz) and peripheral signals such as blood volu...