MACHINE LEARNING
Tiktok Filters
Client: Singapore urban planning campaign #spaceforourdreams (2023)
Creative lead: Corey Cruz
Production lead: Maui Mauricio
Graphic Designers: Dodong Dela Cruz and Conrad Javier
Using a combination of AI and traditional graphic/motion design techniques, I lead a team on the production side that created and animated AI-Generated designs for this Singaporean government advertising campaign.
Welcome to the exciting world of digital art and animation, where the boundaries of creativity are constantly being pushed by innovative technologies. Today, let's explore three groundbreaking tools that are changing the game for artists and animators: ANIMATEDIFF, IPADAPTER, and CONTROLNET.
ANIMATEDIFF: Breathing Life into Static Images
Imagine taking a still image and giving it motion, transforming it into a dynamic, engaging video. That's what ANIMATEDIFF does. This tool is a game-changer for artists looking to venture into animation without the steep learning curve. It's like having a magic wand that animates your static images, opening up new avenues for storytelling and visual expression. With ANIMATEDIFF, the process of animation becomes more accessible and streamlined, allowing even those with limited animation experience to create compelling motion sequences.
IPADAPTER: A Fusion of Text and Image
In the realm of AI-driven art, text-to-image models have been a revelation. But what if you could go further, combining the precision of text prompts with the richness of image inputs? Enter IPADAPTER. This tool revolutionizes how we interact with AI models, enabling more nuanced and detailed image generation. Whether you're refining an existing piece or crafting something new, IPADAPTER brings a level of control and creativity that was previously unattainable. It's particularly exciting for artists who work in multimodal formats, blending text and imagery to create complex, layered compositions.
CONTROLNET: Precision at Your Fingertips
Every artist knows that the devil is in the details. CONTROLNET is about getting those details right. This tool offers unparalleled control over the nuances of AI-generated images and videos. Think of it as the fine brush in your digital art toolkit, allowing you to refine and manipulate images with incredible precision. Whether it's tweaking the depth, smoothing edges, or guiding animations, CONTROLNET ensures that your artistic vision is accurately and beautifully realized.
The Bigger Picture
Together, ANIMATEDIFF, IPADAPTER, and CONTROLNET represent more than just technological advancements; they are empowering tools that expand what's possible in digital art and animation. They offer new ways to explore creativity, whether you're a seasoned professional or just starting out. As these tools continue to evolve, they promise to unlock even more potential for artists and animators, making the digital canvas more vibrant and exciting than ever before.
We've only scratched the surface of what these tools can do. The real magic happens when you start experimenting with them, pushing their limits, and discovering new ways to bring your artistic visions to life. The future of digital art and animation is bright, and it's thrilling to think about what's on the horizon with technologies like ANIMATEDIFF, IPADAPTER, and CONTROLNET leading the way.
CUSTOM FACES
By utilizing stable diffusion and ChatGPT, businesses can bring their marketing and product endorsement strategies to life by creating unique, dynamic characters. These characters can be designed to embody specific brand values, resonate with target audiences, and effectively communicate the product's benefits. With the power of stable diffusion and ChatGPT, businesses have the ability to create compelling, relatable characters that can help drive brand recognition and increase engagement with their target market. Whether it's through interactive chatbots, virtual influencers, or personalized messaging, the possibilities for creative marketing campaigns are endless.
Ceci N'est Pas Un Magritte
Rendered with Deforum Stable Diffusion within Visions of Chaos, processed through RIFE to interpolation x2 frames, then upscaled in Topaz Gigapixel.
Music: Bird's Lament by Moondog
New York Memories
Created using Deforum Stable Diffusion 0.4 within Visions of Chaos.
- Video Input mode @ halfstep 12fps, 896x512
- Each clip has a different prompt in this format: [rough description of actual clip] + [style modifiers]
- Upscaled using Topaz Gigapixel Art & CG mode
- Edited and colored in Premiere
cyberpunk: neo-tokyo
Alternative title: "I plugged in a bunch of anime titles but SD got confused and smooshed them all together"'
Rendered with Deforum Stable Diffusion within Visions of Chaos.
Classic Movie Monsters
Rendered via Stable Diffusion x VoC
an homage to david cronenberg
Generated using Deforum Stable Diffusion 0.3 and upscaled in Topaz Gigapixel
icons of horror
An AI generated video featuring images inspired by the greats of Horror cinema. Video Generated using Stable Diffusion in Google Colab.
Resized (No upscaling) and recolored in Adobe Premiere.
The History of Man in Anime Form
Video Generated using Stable Diffusion in Google Colab. Resized (No upscaling) and recolored in After Effects.
Synthetic Nostalgia
Video Generated using Stable Diffusion in Google Colab. Resized (No upscaling) and recolored in Adobe Premiere.
Cthulhu Comes
Video Generated using Stable Diffusion in Google Colab. Upscaled in Topaz Gigapixel and Recolored in After Effects.