This paper introduces a neural algorithm for artistic style transfer using deep neural networks. The algorithm separates and recombines the content and style of images, enabling the creation of new artistic images. The system uses convolutional neural networks (CNNs) to extract features from images, with higher layers capturing high-level content and lower layers preserving detailed pixel information. A style representation is derived from correlations between features across different layers, capturing texture but not global arrangement. The key finding is that content and style representations in CNNs are separable, allowing independent manipulation to generate new images. The algorithm was tested by combining the content of a photograph with the style of famous artworks, producing images that retain the photograph's content but adopt the artwork's style. The results show that images generated by matching style representations from higher layers are more visually appealing. The method provides a new tool for studying visual perception and neural representations of art. It also offers an algorithmic understanding of how neural networks can independently capture content and style. The work demonstrates that deep neural networks can learn to separate content and style, offering insights into how humans perceive and create artistic imagery. The method is based on the VGG network, a CNN that achieves human-level performance in object recognition. The algorithm uses gradient descent to minimize a loss function that balances content and style reconstruction. The results show that the algorithm can generate images that match both the content of a photograph and the style of an artwork, providing a new approach to artistic style transfer. The method has potential applications in visual perception research, including psychophysics, functional imaging, and electrophysiological studies. The work highlights the power of deep learning in understanding and replicating human visual processing.This paper introduces a neural algorithm for artistic style transfer using deep neural networks. The algorithm separates and recombines the content and style of images, enabling the creation of new artistic images. The system uses convolutional neural networks (CNNs) to extract features from images, with higher layers capturing high-level content and lower layers preserving detailed pixel information. A style representation is derived from correlations between features across different layers, capturing texture but not global arrangement. The key finding is that content and style representations in CNNs are separable, allowing independent manipulation to generate new images. The algorithm was tested by combining the content of a photograph with the style of famous artworks, producing images that retain the photograph's content but adopt the artwork's style. The results show that images generated by matching style representations from higher layers are more visually appealing. The method provides a new tool for studying visual perception and neural representations of art. It also offers an algorithmic understanding of how neural networks can independently capture content and style. The work demonstrates that deep neural networks can learn to separate content and style, offering insights into how humans perceive and create artistic imagery. The method is based on the VGG network, a CNN that achieves human-level performance in object recognition. The algorithm uses gradient descent to minimize a loss function that balances content and style reconstruction. The results show that the algorithm can generate images that match both the content of a photograph and the style of an artwork, providing a new approach to artistic style transfer. The method has potential applications in visual perception research, including psychophysics, functional imaging, and electrophysiological studies. The work highlights the power of deep learning in understanding and replicating human visual processing.