PreciseCam Dataset
收藏资源简介:
PreciseCam数据集由萨拉戈萨大学和Adobe Research联合创建,包含57,380张单视角RGB图像,每张图像均配有文本提示和真实相机参数。该数据集涵盖了广泛的相机参数和多样化的场景,特别适用于解决文本到图像生成中的精确相机控制问题。数据集的创建过程基于用户提供的文本提示和相机参数,通过控制网络(ControlNet)生成图像。该数据集的应用领域主要集中在图像生成和视频生成中,旨在通过精确的相机控制增强生成模型的表达能力,解决传统方法中相机视角控制不足的问题。
The PreciseCam dataset was jointly created by the University of Zaragoza and Adobe Research. It contains 57,380 single-view RGB images, each paired with a text prompt and ground-truth camera parameters. The dataset covers a wide range of camera parameters and diverse scenarios, and is particularly suitable for addressing the problem of precise camera control in text-to-image generation. The dataset is constructed based on user-provided text prompts and camera parameters, with images generated via the ControlNet framework. Its main application fields focus on image and video generation, aiming to enhance the expressive capability of generative models through precise camera control, and solve the problem of insufficient camera view control in traditional methods.




