DiffSynth-Studio/ImagePulseV2-Edit-Merge
ImagePulseV2 Dataset - Multi-Image Merging The ImagePulseV2 dataset is a curated dataset constructed by us for training the Diffusion Templates series of models. It consists of multiple subsets generated using models such as Z-Image-Turbo, Qwen-Image, and Qwen-Image-Edit, based on prompts randomly sampled from DiffusionDB. Open-source code: DiffSynth-Studio Technical report: arXiv Project homepage: GitHub Documentation: English Version, Chinese Version Online demo: ModelScope… See the full description on the dataset page: https://huggingface.co/datasets/DiffSynth-Studio/ImagePulseV2-Edit-Merge.
096
1{2 "edit_prompt_en": "Fuse the figure in a deep blue wide-sleeved hooded robe with an intricately carved golden mask and bare feet running into the foreground center, preserving the robe’s color and adjusting posture to arms outstretched mid-stride; place the figure in a black wide-sleeved hooded robe with the same golden mask and bare feet running slightly behind and to the right at reduced scale, with a subtly trailing motion; replace the background with a desert ruin landscape featuring massive curved stone monoliths and cracked earth, under bright warm lighting to evoke a desolate, epic atmosphere.",3 "edit_prompt_zh": "将身着深蓝色宽袖兜帽长袍、佩戴雕刻金色面具、赤足奔跑的figure置于前景中央，将其衣袍颜色保持为深蓝、姿态微调为双手张开向前奔行；将身着黑色宽袖兜帽长袍、佩戴同款金色面具、赤足奔跑的figure置于其右后方稍远处，缩小比例并调整为略滞后动态姿态；背景替换为荒漠废墟场景，含巨大弧形石质遗迹与干裂沙地，整体光照明亮偏暖，营造苍凉史诗氛围。",4 "mergedimage": [5 "1775952575368407326.png"6 ],7 "seperated image": [8 "1775952575785377782.png",9 "1775952576113366965.png"10 ]11}