Adam & Eve is a two-piece series that experiments with image generation using the Contrastive Language-Image Pre-training (CLIP) model.

The images are generated using a FFT-based generator, guided using CLIP to match the prompts provided in natural language. The matching process is performed using gradient descent, optimizing the FFT parameters to obtain the final images. For each of the images, similar prompts are provided requesting a male (Adam) or female (Eve) body.

Each piece shows, through the eyes of the machine, an abstract image formed by visual features of what constitutes to be a human —male or female— body according to the CLIP model. But, as CLIP has been trained with labels generated by humans, what the machine is showing us is ultimately a reflection of our own perception of human bodies.

Credits

• David Estevez
• Based on Vadim Epstein's early work on CLIP+FTT image generation, that would later become Aphantasia