Visual image reconstruction
In this study, we reconstructed visual images by combining local image bases of multiple scales, whose contrasts were independently decoded from fMRI activity by automatically selecting relevant vox- els and exploiting their correlated patterns. Binary- contrast, 10 × 10-patch images (2^100 possible states) were accurately reconstructed without any image prior on a single trial or volume basis by measuring brain activity only for several hundred random images. Reconstruction was also used to identify the presented image among millions of candidates.
- Yukiyasu Kamitani
- Norihiro Sadato
- Hiroki C. Tanabe
- Yusuke Morito
- Masa-aki Sato
- Okito Yamashita
- Hajime Uchida
- Yoichi Miyawaki
Contact Information:Name: Shuntaro C. Aoki
Acknowledgements and Funding:
The authors thank M. Kawato and K. Toyama for helpful comments; A. Harner and S. Murata for technical assistance; and T. Beck and Y. Yamada for man- uscript editing. This research was supported in part by the SRPBS, MEXT, the NICT-KARC, the Nissan Science Foundation, and the SCOPE, SOUMU.
External Publication Links:Visual image reconstruction from human brain activity using a combination of multiscale local image decoders.
Siemens MAGNETOM Trio A Tim 3T
How to cite this dataset:
In addition to any citation requirements in the dataset summary please use the following to cite this dataset:
This data was obtained from the OpenfMRI database. Its accession number is ds000255
Direct Links to data:
Revision: 1.0.0 Date Set: Jan. 10, 2018, 8:11 p.m.
- Initial release