和传统的服装透视或者透明衣服不同，这个模型更强调在自然光/室内光源的影响下，对身体轮廓的可见性，这种透视往往是真空的、不露点的、非色情的和非湿身的，侧重于与环境的互动和营造温馨的氛围。
Unlike traditional clothing perspective or transparent clothing, this model emphasizes the visibility of body contours under natural light/indoor lighting conditions. This type of perspective is often vacuum-sealed, non-revealing, non-explicit, and non-wet, focusing on the interaction with the environment and creating a cozy atmosphere.
This model is trained based on the original concept of the base model, with the aim of enhancing the effects of certain specific actions, compositions, and poses. In short, it is an entertainment-oriented model~
该模型基于底模原有概念上进行训练，目的是增强某些特定动作/构图/姿态的效果，总而言之，是一个娱乐性的模型~
TAG:
see-through_silhouette,white_shirt
see-through_silhouette, white_dress
example:see-through_silhouette,white_shirt,1girl,who
raining model: Animagine XL V3.1
Inference model: Animagine XL V3.1
Training resolution: 832*1216/1024*1024
Recommended output resolution: 1024*1280
Recommended model weight: 0.8~1
Discord : https://discord.com/invite/CFpyHceF52
Recommended dataset collection and processing tools: https://github.com/deepghs/waifuc
Possibly the most user-friendly dataset processing UI currently in use:https://github.com/Eugeoter/waifuset