| 本文已被：浏览 4371次 下载 7049次 |
码上扫一扫！ |
|
|
| 基于SRN/HMM的非特定人连续手语识别系统 |
|
方高林1, 高文1,2, 陈熙霖1, 王春立3, 马继勇2
|
|
1.哈尔滨工业大学,计算机科学与工程系,黑龙江,哈尔滨,150001;2.中国科学院,计算技术研究所,北京,100080;3.大连理工大学,计算机科学与工程系,辽宁,大连,116023
|
|
| 摘要: |
| 手语识别是通过计算机提供一种有效而准确的机制将手语翻译成文本或语音.目前最新发展水平的手语识别系统在实际应用中应解决非特定人连续手语问题.提出一种将连续手语识别分解成各孤立词识别的分治方法,用于非特定人连续手语识别.把精简循环网(simple recurrent network,简称SRN)作为连续手语的段边界检测器,把SRN分段结果作为隐马可夫模型(hidden Markov models,简称HMM)框架中的状态输入,在HMM框架里使用网格Viterbi算法搜索出一条最佳手语词路径.实验结果表明,该方法的识别效果比单纯使用HMM要好. |
| 关键词: 神经网络 精简循环网络 隐马可夫模型 连续手语识别 非特定人手语识别 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金资助项目(69789301);国家863高科技发展计划资助项目(863-306-ZD03-01-2);中国科学院百人计划资助项目 |
|
| A Signer-Independent Continuous Sign Language Recognition System Based on SRN/HMM |
|
Gao-lin,GAO Wen,CHEN Xi-lin,WANG Chun-li,MA Ji-yong
|
| Abstract: |
| Sign language recognition is to provide an efficient and accurate mechanism to transcribe sign language into text or speech. State-of-the-Art sign language recognition should be able to solve the signer-independent continuous problem for practical applications. In this paper, a divide-and-conquer approach, which takes the problem of continuous CSL (Chinese sign language) recognition as subproblems of isolated CSL recognition, is presented for signer-independent continuous CSL recognition. In the proposed approach, the SRN (simple recurrent network) is used to segment the continuous CSL. The outputs of SRN are regarded as the states of HMM (hidden Markov models) in which the lattice Viterbi algorithm is employed for searching the best word sequence. Experimental results show that SRN/HMM approach has better performance than the standard HMM. |
| Key words: neural network simple recurrent network hidden Markov model continuous sign language recognition signer-independent sign language recognition |
