TY - JOUR AU - Zhao, Sicheng AU - Lin, Chuang AU - Xu, Pengfei AU - Zhao, Sendong AU - Guo, Yuchen AU - Krishna, Ravi AU - Ding, Guiguang AU - Keutzer, Kurt PY - 2019/07/17 Y2 - 2024/03/28 TI - CycleEmotionGAN: Emotional Semantic Consistency Preserved CycleGAN for Adapting Image Emotions JF - Proceedings of the AAAI Conference on Artificial Intelligence JA - AAAI VL - 33 IS - 01 SE - AAAI Technical Track: Humans and AI DO - 10.1609/aaai.v33i01.33012620 UR - https://ojs.aaai.org/index.php/AAAI/article/view/4110 SP - 2620-2627 AB - <p>Deep neural networks excel at learning from large-scale labeled training data, but cannot well generalize the learned knowledge to new domains or datasets. Domain adaptation studies how to transfer models trained on one labeled source domain to another sparsely labeled or unlabeled target domain. In this paper, we investigate the unsupervised domain adaptation (UDA) problem in image emotion classification. Specifically, we develop a novel cycle-consistent adversarial model, termed CycleEmotionGAN, by enforcing emotional semantic consistency while adapting images cycleconsistently. By alternately optimizing the CycleGAN loss, the emotional semantic consistency loss, and the target classification loss, CycleEmotionGAN can adapt source domain images to have similar distributions to the target domain without using aligned image pairs. Simultaneously, the annotation information of the source images is preserved. Extensive experiments are conducted on the ArtPhoto and FI datasets, and the results demonstrate that CycleEmotionGAN significantly outperforms the state-of-the-art UDA approaches.</p> ER -