The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Corrective learning is a technique that applies classification methods for automatically detecting and correcting systematic segmentation errors produced by existing segmentation methods with respect to some gold standard (manual) segmentation. To allow corrective learning more effectively correct errors that require non-local contextual information to capture, we extend the corrective learning technique...
Shape priors have been successfully used in challenging biomedical imaging problems. However when the shape distribution involves multiple shape classes, leading to a multimodal shape density, effective use of shape priors in segmentation becomes more challenging. In such scenarios, knowing the class of the shape can aid the segmentation process, which is of course unknown a priori. In this paper,...
Model M, an exponential class-based language model, and neural network language models (NNLM's) have outperformed word n-gram language models over a wide range of tasks. However, these gains come at the cost of vastly increased computation when calculating word probabilities. For both models, the bulk of this computation involves evaluating the softmax function over a large word or class vocabulary...
We present a contextual spoken language understanding (contextual SLU) method using Recurrent Neural Networks (RNNs). Previous work has shown that context information, specifically the previously estimated domain assignment, is helpful for domain identification. We further show that other context information such as the previously estimated intent and slot labels are useful for both intent classification...
In this paper we introduce a novel low dimensional method to perform topic detection and classification in Twitter. The proposed method first employs Joint Complexity to perform topic detection. Then, based on the nature of the data, we apply the theory of Compressive Sensing to perform topic classification by recovering an indicator vector, while reducing significantly the amount of information from...
Channel state information at the transmitter (CSIT) is essential for frequency-division duplexing (FDD) massive multiple-input multiple-output (MIMO) systems, but conventional solutions involve overwhelming overhead both for downlink channel training and uplink channel feedback. In this paper, we propose a joint CSIT acquisition scheme based on low-rank matrix recovery to reduce the overhead. Particularly,...
The thyroid gland is an important endocrine organ. For a variety of clinical applications, a system for automated segmentation of the thyroid is desirable. Thyroid segmentation is challenging due to the inhomogeneous nature of the thyroid and the surrounding structures which have similar intensities. In this paper, we propose a fully automated method for thyroid detection and segmentation on CT scans...
Based on the recently proposed speech pre-processing front-end with deep neural networks (DNNs), we first investigate different feature mapping directly from noisy speech via DNN for robust speech recognition. Next, we propose to jointly train a single DNN for both feature mapping and acoustic modeling. In the end, we show that the word error rate (WER) of the jointly trained system could be significantly...
In this paper, we consider multi-sensor classification when there is a large number of unlabeled samples. The problem is formulated under the multi-view learning framework and a Consensus-based Multi-View Maximum Entropy Discrimination (CMV-MED) algorithm is proposed. By iteratively maximizing the stochastic agreement between multiple classifiers on the unlabeled dataset, the algorithm simultaneously...
This article proposes and evaluates a Gaussian Mixture Model (GMM) represented as the last layer of a Deep Neural Network (DNN) architecture and jointly optimized with all previous layers using Asynchronous Stochastic Gradient Descent (ASGD). The resulting “Deep GMM” architecture was investigated with special attention to the following issues: (1) The extent to which joint optimization improves over...
In the hybrid approach, neural network output directly serves as hidden Markov model (HMM) state posterior probability estimates. In contrast to this, in the tandem approach neural network output is used as input features to improve classic Gaussian mixture model (GMM) based emission probability estimates. This paper shows that GMM can be easily integrated into the deep neural network framework. By...
In this paper, we present methods in deep multimodal learning for fusing speech and visual modalities for Audio-Visual Automatic Speech Recognition (AV-ASR). First, we study an approach where uni-modal deep networks are trained separately and their final hidden layers fused to obtain a joint feature space in which another deep network is built. While the audio network alone achieves a phone error...
Information theoretic image and volume registration is currently of interest as a method for multi-modal alignment. It has been suggested that it is useful to incorporate information obtained from previous registrations into these methods to improve future registration performance. In this paper we examine how this can be done when using a graph theoretic estimator of entropy. Our main contribution...
Determining a systems design, analysis or approach to be of high or low quality remains a subjective assessment. Our field requires the ability to objectively grade the quality of a systems approach in advance of implementation and then correlate that assessment with outcomes.
In this paper, an sEMG-driven musculoskeletal model of human shoulder and elbow joints is built based on time delay neural network (TDNN). Six principal muscles of the upper arm and forearm are included, and the experiment was conducted under isometric contractions with the aid of a planar haptic interface. Both force amplitude and direction were regulated continuously, and the experiment results...
As sequelae of stroke, hemiplegia is a typical symptom, where paralysis occurs in the half the body because of brain damage. In this study, we have developed selectable constraint mechanism for hemiplegic upper limb training. By using this mechanism, it became possible to separate synergic movement while flexion-extension training of shoulder and elbow by constraining each individual joints. As the...
We developed KneeRobo, which replicates knee joint disorders experienced by patients, to enable students studying to become physical or occupational therapists to gain practical training/testing experience virtually. We also developed a control algorithm that enables KneeRobo to realize involuntary internal/external rotation and abduction/adduction during knee flexion and extension. However, in the...
In recent years, the population of elderly people is increasing rapidly. Rehabilitation training systems using robotics and virtual reality technologies, therefore, attracts attention. This paper introduces the development of an upper limb rehabilitation robot with guidance control. This study investigates how the motor learning effect improves with low stiffness guidance control based on pneumatic...
A golf swing training system has been designed to allow users to improve their skill through careful guidance along an ideal golf swing trajectory. The training system is designed as a 6-DOF robot manipulator which controls the position and orientation of the golf club grip point. The manipulator's position and velocity kinematics are described using the DH coordinate frame method, from which a closed-form...
In this paper, We present a Self-Physical Rehabilitation System (SPRS) that assists patients to rehab correctly at home by themselves without physical therapists for solving the problem of access to medical service. SPRS uses KINECT to detect movement of patient's body while doing self-rehab and processing of data from motion detection of patients by KINECT to determine accuracy of gesture. This software...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.