A CNN-transformer hybrid approach for decoding visual neural activity into text.
Jiang ZhangChen LiGanwanming LiuMin MinChong WangJiyi LiYuting WangHongmei YanZhentao ZuoWei HuangHuafu ChenPublished in: Comput. Methods Programs Biomed. (2022)
Keyphrases
- network architecture
- visual information
- visual processing
- web images
- database
- primary visual cortex
- convolutional neural network
- cellular neural networks
- neural network
- semantic content
- biologically motivated
- text mining
- keywords
- visual features
- video search
- high level
- information retrieval
- video segments
- free text
- text documents
- power system
- low level
- automatically extracted
- bio inspired
- visual cortex
- semantic information
- image quality
- fuzzy logic
- color images
- neural mechanisms