Publications by authors named "Yu-Hui Zuo"
Article Synopsis
- - A new method for object classification using a single-pixel neural network is introduced, allowing direct classification from bucket measurements without the need for full image reconstruction.
- - This method achieves a high classification accuracy of 94.23% using only 16 measurements, significantly below the Nyquist limit, and employs parallel computing to speed up data processing.
- - The approach improves performance over existing techniques for both binary and grayscale images in challenging conditions, making it potentially useful in areas like remote sensing and military defense.
