TripleView is a LoRA trained on approximately 300 character sheet samples, with a strong focus on generating characters in three main views:
If this model was helpful to you in any way, please consider helping me continue creating more model for the price of a coffee.
Front · Side · Back
Training Focus
TripleView was trained primarily on human characters. It may also work with cartoon, anime, or stylized characters, but these styles were not the main focus of the training dataset, so results may be less consistent.
Its purpose is to help convert a character into a clean character sheet while maintaining a pleasant level of consistency across clothing, proportions, appearance, and overall character identity.
This LoRA is not perfect and may still produce variations in anatomy, accessories, facial details, clothing, or pose alignment. However, it generally preserves character consistency in a visually pleasing way, making it useful for character design, reference sheets, dataset creation, and further refinement.
TripleView was trained primarily to be used as part of a CharacterID workflow for LTX, where the generated three-view sheet can later serve as reference material to improve character identity preservation.
Trigger
Flux Klein Prompt: "Convert the character in the image to a Character Sheet showing front, side and back full body views"
Krea 2 Prompt: "Convert the character in the image to a Character Sheet showing a face close-up, front full body, side full body and back full body views"
After the trigger, you can also describe the type of clothing you want the character to wear.
For example, you can add outfit details right after the trigger to specify the character’s clothes, style, or overall look.
Recommended for
Character sheets
Three-view turnarounds
Front / side / back references
Character dataset preparation
CharacterID workflows
LTX pipelines
Description
This version is highly experimental, and the text generation is still far from perfect. I tested many different configurations, and the best results came from using the LCM sampler with the Simple scheduler.
You can also try increasing the resolution to 2048 × 1365, 3072 × 2048, or 4096 × 2736. The default resolution is 1536 × 1024, but higher resolutions can improve text quality slightly. Even so, don't expect perfect results yet.
Consistency is not 100%, although it tends to be a bit better when using LCM. Overall, I think this was a successful experiment, and I'll continue improving it in future updates.
The prompt must follow the specific format used in the workflow. You can write it manually if you want, but using the included VLM to generate the prompt usually produces better and more consistent results, especially when it comes to preserving the expected structure.
The model is also quite versatile. It was trained using my previous 4 Views model as its foundation, so you're not limited to the included prompt. Feel free to customize it or even reuse the prompt from the original 4 Views workflow to generate different styles and layouts of character sheets. Experimenting with different prompt structures can produce surprisingly good results.
If you discover a better way to improve the text generation, feel free to leave a comment and share your findings.