This LoRA is designed for female subjects (18+) in standing poses, typically framed as a portrait, and often partially or fully nude. The defining characteristic of this model is that the subject is always positioned behind some object—a tree, leaves, a wall, partially submerged in water, behind curtains, or similar. In other words, the poser is partially concealed behind an element that adds depth and intrigue to the composition.This is an erotic-type LoRA with a strong compositional focus. It functions as a detail LoRA, adding accent and atmosphere to the scene rather than dictating a specific anatomical pose.<hr />Subject &amp; Pose DescriptionIn dataset subject is depicted standing, usually facing the camera directly—there are no significant twists, turns, or rear-view shots. The pose is straightforward, with the emphasis placed on the interplay between the subject and the object partially obscuring her.In many cases, the accent falls on the exposed breast, while the genital area remains minimally visible due to the framing and the concealing object. The overall effect is suggestive and artistic, with the object acting as a compositional framing device.<hr />Strength0.5 - 1.0: The effect begins to activate; subtle introduction of the concealing object and compositional accent.1.5 – 2.5: Stronger, more pronounced effect; the object framing becomes a dominant compositional element, enhancing the erotic and artistic quality.<hr />Image Format RecommendationThe vast majority of training images are in vertical orientation. Therefore, it is highly recommended to generate images in portrait mode (9:16 aspect ratio) for optimal composition and framing.<hr />Training &amp; Captioning Methodology<ul><li>Detailed Analysis: Each training image underwent thorough analysis and captioning with an exclusive focus on how the poser is positioned behind the object. This includes the type of object, its texture and visual details, how it partially conceals the figure, and the resulting compositional effect. The interplay between the subject and the concealing element was the primary consideration - not the garment or body parts in isolation.</li></ul><hr />Prompting RecommendationThis model was trained on natural language captions. For best results, use descriptive full sentences rather than comma-separated tag lists.Suggested concepts:<ul><li><code>standing behind a tree</code></li><li><code>partially hidden behind leaves</code></li><li><code>behind a wall</code></li><li><code>partially submerged</code></li><li><code>behind curtains</code></li><li><code>peeking from behind an object</code></li><li><code>standing nude behind [object]</code></li><li><code>portrait composition</code></li></ul><hr />License &amp; Usage<ul><li>You are free to use this model in your projects, including merging or mixing it with other models.</li><li>If you include this LoRA in any merge or mix, please give proper credit.</li></ul><hr />Support the CreatorIf you enjoy my work and want to help cover the electricity costs required to train and refine these models, please consider supporting me:Address: <a target="_blank" rel="ugc" href="https://ko-fi.com/weirddreamer">https://ko-fi.com/weirddreamer</a> (Donations are always appreciated!)

Hidden Allure (Krea2 & Moss)

Dataset Size:108 high-quality, manually curated imagesData Processing:Images were meticulously filtered using advanced duplicate removal tools to ensure maximum variety and uniqueness. Additionally, the dataset underwent an experimental terminology simplification process, making it easier to describe the desired results in plain, everyday language in the final outputs.Captions:Tagging was performed using Ornith-1.5

Files with SHA256: cf6fe59f003ad89cf0e04d8120057a61b685401478e71821df5a55aaf85860cb