Copy Motion From One to Another: Fake Motion Video Generation
Authors: Zhenguang Liu, Sifan Wu, Chejian Xu, Xiang Wang, Lei Zhu, Shuang Wu, Fuli Feng
IJCAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments show that our method is able to generate realistic target person videos, faithfully copying complex motions from a source person. Our code and datasets are released at https://github.com/Sifann/Fake Motion |
| Researcher Affiliation | Academia | Zhenguang Liu1,2 , Sifan Wu2 , Chejian Xu1 , Xiang Wang3 , Lei Zhu4 , Shuang Wu5 and Fuli Feng6 1Zhejiang University 2Zhejiang Gongshang University 3National University of Singapore 4Shandong Normal Unversity 5Nanyang Technological University 6University of Science and Technology of China |
| Pseudocode | No | The paper includes architectural diagrams (Fig. 1, Fig. 2) and mathematical equations, but no structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Our code and datasets are released at https://github.com/Sifann/Fake Motion |
| Open Datasets | Yes | Experiments are conducted on two benchmark datasets, i PER [Liu et al., 2019a] and Complex Motion. ...Our code and datasets are released at https://github.com/Sifann/Fake Motion |
| Dataset Splits | No | The paper does not explicitly provide training/test/validation dataset splits or cross-validation methodology. It mentions using 'benchmark datasets' and 'mini-batch of 10 for 120 epochs' but no specific split details for validation. |
| Hardware Specification | Yes | We train our model with a mini-batch of 10 for 120 epochs on a Nvidia RTX 2080-Ti GPU. |
| Software Dependencies | No | The paper mentions 'Open Pose' and 'Mask-RCNN' but does not provide specific version numbers for these or other software dependencies. |
| Experiment Setup | Yes | During training, all the images are resized to 512 512. We train our model with a mini-batch of 10 for 120 epochs...The initial learning rate is set to 1e 4. We employ the Adam optimizer with β1 = 0.9 and β2 = 0.999. |