搜索结果: 76-90 共查到“知识库 模式识别”相关记录1542条 . 查询时间(2.875 秒)
基于多种上下文结构与线性融合的特定敏感视频识别
特定敏感视频 多种上下文结构 分类融合 线性依赖模型
2017/1/9
本文中特定敏感视频是指恐怖和暴力视频,现有的特定敏感视频识别算法或是忽略了视频的多种上下文结构信息;或是忽略了各种特征间潜在的依赖关系.因此,本文提出了一种基于多种上下文结构与线性融合的特定敏感视频识别方法,首先针对某种视频提取多种有效特征,并获取镜头间的上下文结构信息;然后,在每一个特征空间中利用上下文结构训练一个SVM分类器;最后,获取不同特征间的依赖关系,采用线性依赖模型融合多个分类器的结果...
Automated Detection and Counting of Pedestrians on an Urban Roadside
Pedestrian detection Pedestrian counting Transportation planning Computer vision Autoscope Histogram of oriented gradients
2014/12/8
This thesis implements an automated system that counts pedestrians with 85% accuracy. Two approaches have been considered and evaluated in terms of count accuracy, cost and ease of deployment. The fir...
Indoor Navigation System for the Visually Impaired with User-centric Graph Representation and Vision Detection Assistance
Indoor Navigation System Visually Impaired User-centric Graph Representation Vision Detection Assistance
2014/12/8
Independent navigation through unfamiliar indoor spaces is beset with barriers for the visually impaired. Hence, this issue impairs their independence, self-respect and self-reliance. In this thesis I...
基于随机PRI压缩感知雷达的速度假目标识别方法
随机重复间隔信号 压缩感知 假目标识别 动目标检测
2017/1/16
当存在有源速度假目标干扰时,随机脉冲重复间隔(Random Pulse Repetition Interval,RPRI)压缩感知雷达的回波稀疏性将被破坏,其目标检测与测速性能无法正常发挥.考虑到干扰机通常无法保证在相位调制时实时准确地更新雷达随机脉冲重复间隔,这将导致其生成的虚假目标信号与真实目标信号在不同的字典上稀疏.本文利用这一特性,在压缩感知框架下提出速度假目标识别算法,该算法通过建立联合...
Limited Labels for Unlimited Data: Active Learning for Speaker Recognition
speaker recognition i-vectors active learning
2014/11/27
In this paper, we attempt to quantify the amount of labeled data necessary to build a state-of-the-art speaker recognition system. We begin by using i-vectors and the cosine similarity metric to repre...
Non-Negative Factor Analysis for Gaussian Mixture Model Weight Adaptation for Language and Dialect Recognition
Non-negative factor analysis model adaptation Gaussian mixture model weight dialect recognition language recognition
2014/11/27
Non-Negative Factor Analysis for Gaussian Mixture Model Weight Adaptation for Language and Dialect Recognition.
Speech Feature Denoising and Dereverberation via Deep Autoencoders for Noisy Reverberant Speech Recognition
robust speech recognition feature denoising denoising autoencoder deep neural network
2014/11/27
Denoising autoencoders (DAs) have shown success in generating robust features for images, but there has been limited work in applying DAs for speech. In this paper we present a deep denoising autoenco...
SPOKEN COMMAND OF LARGE MOBILE ROBOTS IN OUTDOOR ENVIRONMENTS
Human-robot interaction real-time speech recognition voice activity detection modulation frequency
2014/11/27
We describe a speech system for commanding robots in human-occupied outdoor military supply depots. To op erate in such environments, the robots must be as easy to interact with as are humans, i.e. th...
A Back-off Discriminative Acoustic Model for Automatic Speech Recognition
context-dependent acoustic modeling back-off acoustic models discriminative training
2014/11/27
In this paper we propose a back-off discriminative acoustic model for Automatic Speech Recognition (ASR). We use a set of broad phonetic classes to divide the classification problem originating from c...
Research Developments and Directions in Speech Recognition and Understanding, Part 1
Research Developments Speech Recognition Understanding
2014/11/27
To advance research, it isimportant to identify prom-ising future research direc-tions, especially those thathave not been adequately pursued or funded in the past. The working group producing this ar...
LANGUAGE MODEL PARAMETER ESTIMATION USING USER TRANSCRIPTIONS
speech recognition, language modeling, adaptation
2014/11/27
In limited data domains, many effective language modeling techniques construct models with parameters to be estimated on an in-domain development set. However, in some domains, no such data exist beyo...
Robust Speaker Recognition in Noisy Conditions
Missing-feature theory multicondition training noise compensation noise modeling speaker recognition
2014/11/27
This paper investigates the problem of speaker identification and verification in noisy conditions, assuming that speech signals are corrupted by environmental noise, but knowledge about the noise cha...
肌电模糊熵特征的加权核FDA跌倒识别
跌到检测 表面肌电信号 模糊熵 加权核Fisher线性判别 模式识别
2016/12/29
针对跌倒常常对老年人的健康构成严重危害的问题.本文设计了一种基于肌电信号的跌倒检测方法,首先提取腓肠肌和股外侧肌的sEMG的模糊熵特征作为特征向量,然后,针对日常活动动作类(Activities of Daily Life,ADL)的数目远多于跌倒类导致的数据集不平衡的问题,提出了加权核Fisher线性判别方法,采用相应的平衡参数来调节样本核矩阵,最终,将跌倒与行走、蹲下和坐下辨识出来.实验结果表...
一种基于高阶累积量的正交空时分组码盲识别方法
空时分组码 高阶累积量 信号识别
2016/12/29
针对全盲条件下正交空时分组码(OSTBC)的正交性识别问题,提出一种基于高阶累积量的OSTBC识别方法.通过对接收信号白化,在不需要信道估计的情况下,消除信道对识别结果的干扰,既减少了算法复杂度,又提高了低信噪比条件下OSTBC的识别概率;考虑实际系统中噪声对识别结果的影响,利用高阶累积量(阶数大于2)的去噪声特性,消除噪声对于STBC正交特性识别的影响;分析了两种四阶累积量的OSTBC识别特性,...