gen AI
Experimentos y proyectos creados utilizando herramientas de inteligencia artificial generativa
YEAR: 2024 CREATED WITH COMFYUI, FLUX.1-DEV AND KLING 1.6
SHOWREEL 2025 (IA)
Un showreel que muestra mi trabajo reciente con IA generativa.
YEAR: 2024 CREATED WITH COMFYUI, FLUX.1-DEV AND KLING 1.6
"PRODUCT ANIMATION"
A test of product animation with Flux and Kling 1.6. While Flux successfully handles text in most cases, for Kling, it remains a challenge. If this were client work, you’d need to do some cleanup and text/logo replacement. However, I imagine that video generation models of even better quality and consistency than Kling 1.6 will very soon be available in ComfyUI, with its incredible flexibility and control. The era of rendered content is inevitably coming to an end, giving way to generated content.
YEAR: 2024 CREATED WITH COMFYUI, FLUX.1-DEV AND KLING 1.6
"NIKE JORDAN"
Another GenAI experiment—this time with Nike Jordan! 🏀✨
I used a trained LoRa model for the product, generated still images in Flux-dev with ComfyUI, and created several video generations using Kling 1.6. After that, I combined a few shots, added a touch of motion blur to smooth out the usual warping you get in dynamic shots, and finished it off with music and SFX.
YEAR: 2024 CREATED WITH COMFYUI, FLUX.1-DEV AND KLING 1.6
"LA DANCE DES BAUES ROUGES"
Images generated with Flux in Comfy UI using Redux for color and style consistency, and depth ControlNet for composition guidance.
Depth maps created in Blender and Cinema 4D.
Video generated in Kling 1.6.
YEAR: 2024 CREATED WITH COMFYUI, FLUX.1-DEV AND KLING 1.6
"HONEY VALLEY"
Lately, I’ve been exploring ways to integrate GenAI into my production workflows. While certain types of shots remain challenging—or even impossible—due to consistency issues or hallucinations, there are specific content types where AI truly excels.
One area I’ve been experimenting with recently is viscous liquid shots, which are a staple in TV commercials. Traditionally, creating realistic fluid simulations in 3D applications like Houdini or Cinema 4D is both time-consuming and resource-intensive.
With GenAI, however, I’ve discovered an incredible time-saving alternative that delivers impressive results for these kinds of visuals.
It’s exciting to see how this technology is transforming aspects of production, and I’m looking forward to exploring its potential further
YEAR: 2024 CREATED WITH COMFYUI, FLUX.1-DEV AND KLING 1.6
"CHOCOLATE CHARM"
YEAR: 2024 FOOTAGE GENERATED WITH STABLE DIFFUSION AND LUMA AI | COMPOSITED IN AFTER EFFECTS
