Automated Product Rendering Workflow Needed
Budget / SalaryHourly project
TypeFreelance project
LocationRemote
Posted2 hours ago
I need a complete ComfyUI workflow that lets a store-owner walk in, drop one garment image (shirt or pants), make a few quick selections, and instantly receive three 1 K renders ready for the product page.
Core flow I have in mind
• Ten pre-defined fashion models (5 women, 5 men).
• Sticky Pose control: the user chooses one option from a curated pose list; no free-form entry needed.
• Background comes from a text prompt field and should blend cleanly with the figure.
• Aspect ratio switcher offering only 1:1 or 3:4 or 9:16 or 2:3
• Output: three images per run, consistent lighting and colour, ready for e-commerce.
Everything must run inside ComfyUI—no external scripts except what can be called through nodes. If you need ControlNet, LoRA, or in-painting nodes, wire them in so they work out of the box; the fewer manual steps, the better. I will supply the garment cut-outs as PNGs, so the graph must accept a single file upload and composite it correctly onto the chosen model and pose.
Deliverables
1. The .json workflow file with all nodes configured.
2. Any custom checkpoints, LoRAs, or control images referenced, packaged or linked.
3. A short README so I can drop everything into my ComfyUI folder and reproduce your results.
1. Do you already have the ten model images or references ready to share?
2. Roughly how many poses should the curated pose list include?
3. Do you have a preferred base… more
Hi, could you confirm whether the supplied garment PNGs will already be pre-masked (transparent background and aligned to a fixed garment canvas) or should the workflow include an automatic… more
Core flow I have in mind
• Ten pre-defined fashion models (5 women, 5 men).
• Sticky Pose control: the user chooses one option from a curated pose list; no free-form entry needed.
• Background comes from a text prompt field and should blend cleanly with the figure.
• Aspect ratio switcher offering only 1:1 or 3:4 or 9:16 or 2:3
• Output: three images per run, consistent lighting and colour, ready for e-commerce.
Everything must run inside ComfyUI—no external scripts except what can be called through nodes. If you need ControlNet, LoRA, or in-painting nodes, wire them in so they work out of the box; the fewer manual steps, the better. I will supply the garment cut-outs as PNGs, so the graph must accept a single file upload and composite it correctly onto the chosen model and pose.
Deliverables
1. The .json workflow file with all nodes configured.
2. Any custom checkpoints, LoRAs, or control images referenced, packaged or linked.
3. A short README so I can drop everything into my ComfyUI folder and reproduce your results.
1. Do you already have the ten model images or references ready to share?
2. Roughly how many poses should the curated pose list include?
3. Do you have a preferred base… more
Hi, could you confirm whether the supplied garment PNGs will already be pre-masked (transparent background and aligned to a fixed garment canvas) or should the workflow include an automatic… more
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.