ReMix: Towards Image-to-Image Translation with Limited Data | |
Cao, Jie1,2; Hou, Luanxuan1,2; Yang, Ming-Hsuan3; He, Ran1,2![]() ![]() | |
2021 | |
会议名称 | IEEE Conference on Computer Vision and Pattern Recognition |
会议日期 | 2021年6月19日 – 2021年6月25日 |
会议地点 | 美国田纳西州纳什维尔 |
摘要 | Image-to-image (I2I) translation methods based on generative adversarial networks (GANs) typically suffer from overfitting when limited training data is available. In this work, we propose a data augmentation method (ReMix) to tackle this issue. We interpolate training samples at the feature level and propose a novel content loss based on the perceptual relations among samples. The generator learns to translate the in-between samples rather than memorizing the training set, and thereby forces the discriminator to generalize. The proposed approach effectively reduces the ambiguity of generation and renders content-preserving results. The ReMix method can be easily incorporated into existing GAN models with minor modifications. Experimental results on numerous tasks demonstrate that GAN models equipped with the ReMix method achieve significant improvements. |
收录类别 | EI |
语种 | 英语 |
是否为代表性论文 | 否 |
七大方向——子方向分类 | 多模态智能 |
国重实验室规划方向分类 | 小样本高噪声数据学习 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/44726 |
专题 | 模式识别实验室 |
通讯作者 | He, Ran |
作者单位 | 1.智能感知与计算研究中心 2.中国科学院大学 3.加州大学默塞德分校 |
推荐引用方式 GB/T 7714 | Cao, Jie,Hou, Luanxuan,Yang, Ming-Hsuan,et al. ReMix: Towards Image-to-Image Translation with Limited Data[C],2021. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
00633.pdf(4848KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论