辨识阅读眼动注视点参数的级联神经网络
A Cascade Neural Network for Identifying Fixation Parameters of Reading Eye-Movements
-
摘要: 提出了一种由三层感知机和单层线性网络构成的级联神经网络,能用于自动辨识眼 动实验数据中的注视停顿并确定其时程.计算机仿真结果表明,该网络经过学习训练后能快 速、准确地实时处理眼动实验数据,相对误差小于0.2％.Abstract: A cascade neural network, which is constructed with a three-layer perceptron and a one-layer linear network, is provided for identifying fixation pauses and fixation durations in eye movements. Simulation results show that computation of fixation parameters of eye-movement data with this neural network is accurate and fast.
-
Key words:
- Neural network /
- pattern identification /
- reading eye movement /
- fixation pause /
- fixation duration
-
计量
- 文章访问数: 2236
- HTML全文浏览量: 159
- PDF下载量: 948
- 被引次数: 0
下载:
