- A script that applies the AdaIN style transfer method to VOC format object detection datasets.
- 本项目是使用 AdaIN 对 VOC 格式的目标检测数据集进行风格迁移的脚本，可以同时生成风格迁移后的图像和对应的标签文件。
- 本项目在 stylize-datasets 基础上进行改进，在原项目的基础上新增了标签文件的生成；
- 由于风格迁移后的图像中物体的形状和位置信息基本不改变，所以重新生成的标签文件中的
object信息不发生变化，仅在filename和path中改变新图像生成的名称和位置。
stylize a dataset, run python stylize.py. Arguments like below:
--content-dir <CONTENT>the top-level directory of the content image dataset (mandatory)--style-dir <STLYE>the top-level directory of the style images (mandatory)--output-dir <OUTPUT>the directory where the stylized dataset will be stored (optional, default: output/)--num-styles <N>number of stylizations to create for each content image (optional, default: 1)--alpha <A>Weight that controls the strength of stylization, should be between 0 and 1 (optional, default: 1)--extensions <EX0> <EX1> ...list of image extensions to scan style and content directory for (optional, default: png, jpeg, jpg). Note: this is case sensitive, --extensions jpg will not scan for files ending on .JPG. Image types must be compatible with PIL's Image.open() (Documentation)--content-size <N>Minimum size for content images, resulting in scaling of the shorter side of the content image to N (optional, default: 0). Set this to 0 to keep the original image dimensions.--style-size <N>Minimum size for style images, resulting in scaling of the shorter side of the style image to N (optional, default: 512). Set this to 0 to keep the original image dimensions (for large style images, this will result in high (GPU) memory consumption).--crop <N>Size for the center crop applied to the content image in order to create a squared image (optional, default 0). Setting this to 0 will disable the cropping.--src-label-path <CONTENT LABEL>the top-level directory of the xml label for content image dataset--out-label-path <STLYE LABEL>the top-level directory of the xml label for style image dataset