WE BEGAN FROM A SIMPLE BELIEF: IN THE DEPTH OF WINTER, TRUE LUXURY IS SOFTNESS YOU CAN FEEL. SOFT WINTER COLLECTION PROECT FRAMES GENTLE, DREAM-LIKE SILHOUETTES AGAINST SHARP ALPINE LANDSCAPES—PINK FUR, KNIT TEXTURES, AND QUIET LIGHT WRAPPING THE COLD IN WARMTH. THESE IMAGES TURN FROZEN AIR, DISTANCE, AND SNOW INTO TENDER MOMENTS THAT LINGER LONG AFTER WINTER PASSES.
Client
Senseya
DELIVERABLES
EDITORAIL VISUALS BRANDING
Year
2025
Role
Creative Direction
Tools
MIDJOURNEY, SEEDREAM, KLING AI
Soft Winter Collection
I led the creative direction and visual system for this Soft Winter editorial, using only generative AI tools to shape the brand world. Starting from a palette of blush fur, knit textures, and pale alpine light, The result is a seasonal campaign toolkit: a set of stills and sequences that a brand could adapt across lookbooks, social launches, in-store screens, and motion teasers, all anchored in one clear warm winter story.
I iterated in Midjourney, Higgsfield AI, Seedream, Kling AI, and VEO to lock in a consistent character, wardrobe, and camera language across every frame. I treated prompts like a styling and lighting kit—tuning pose, fabric movement, snowfall density, and time of day until the images felt soft, cinematic, and cohesive.
Process: Generative Workflow System
Notion AI TOOLS & Workflow REsearch ARCHIVING
I. Transformation Pipeline
From static image to motion sequence
This workflow explores how generative AI can be directed as a multi-step transformation system rather than a single output tool.Starting from a base image,
I iteratively modify structure, environment, and motion through chained prompts and models.
Workflow
• Outfit transformation using prompt-based image editing
• Background reconstruction while preserving subject identity
• Motion generation through video models (Kling)
• Final sequencing into continuous movement
AI Tools
Tools: Midjourney · Seedream · Kling · Figmaweave
Focus: Transformation / Motion / Iterative control
Insight
This process reframes AI as a system for controlling transitions across states, not just generating isolated images.
II. Visual System
Camera, framing, and consistency
This exploration focuses on how AI can maintain visual coherence across different camera perspectives and framing conditions.
By systematically testing shot types, I establish a controlled visual language across outputs.
Exploration
Eye-level and high-angle framing
Tight close-ups vs full-body compositions
Side profiles and three-quarter views
Consistency of lighting, styling, and identity across shots
Insight
Rather than generating individual images, this approach builds a repeatable visual system that can scale across campaigns, lookbooks, and motion.