REINVENTING BLACK-AND-WHITE PHOTOGRAPHY WITH AI FILTERS

Abstract [English]

Reinvention of black-and-white (B&W) photography by artificial intelligence (AI) is an example of the paradigm shift in the concept of tone expression, texture representation, and emotion in monochrome photography. The classical B and W processes, determined by the chemistry of films and optical exposure, placed more importance on the tonal gradation, the shadow-highlight relationship, and the contrast relationship. These artistic nuances, however, were prone to being lost with the shift to the digital workflows because of the linear desaturation and channel-based conversions. The suggested research presents an AI-based framework redefining creative and technical limits of digital monochrome photography based on convolutional neural networks (CNNs), generative adversarial networks (GANs), and diffusion-based models. It can be explained by the following methodology: The resulting diversified in terms of genres dataset, such as portraits, landscapes, architecture, and abstract textures, is prepared and then adaptive tone-filtering filters and contrast-conscious CNN modules are designed. A series of steps in tonal reconstruction pipeline also guarantees region-based luminance adjustment, noise reduction and preservation of detail. Quantitative values like PSNR, SSIM, and LPIPS alongside subjective evaluations reveal that every method has shown great enhancement in the tonal depth, clarity, and aesthetic realism over classical and current digital conversion. In addition to objective fidelity, the framework increases the emotional evocation of images, re-creating the classical richness of analog B&W and adding the current computational accuracy.