One-shot Face Reenactment Using Appearance Adaptive Normalization
Authors: Guangming Yao, Yi Yuan, Tianjia Shao, Shuang Li, Shanqi Liu, Yong Liu, Mengmeng Wang, Kun Zhou3172-3180
AAAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive quantitative and qualitative experiments demonstrate the significant efficacy of our model compared with prior one-shot methods. |
| Researcher Affiliation | Collaboration | 1 Net Ease Fuxi AI Lab 2 State Key Lab of CAD&CG, Zhejiang University 3 School of Computer Science and Technology, Beijing Institute of Technology 4 Institute of Cyber-Systems and Control, Zhejiang University |
| Pseudocode | No | The paper describes its methodology and model architecture using text and diagrams (e.g., Figure 2, 3, 4) but does not include any structured pseudocode or algorithm blocks. |
| Open Source Code | No | The paper does not provide any explicit statement about making its source code publicly available, nor does it include a link to a code repository. |
| Open Datasets | Yes | Both the Face Forensics++ (R ossler et al. 2019) and Celeb DF (Li et al. 2020) datasets are used for quantitative and qualitative evaluation. |
| Dataset Splits | No | The paper mentions a 'training stage' and 'testing set' but does not provide specific details on how the datasets were split into training, validation, and test sets, such as exact percentages, sample counts, or the methodology for partitioning the data. |
| Hardware Specification | No | The paper does not provide specific hardware details such as exact GPU/CPU models, processor types, or memory amounts used for running its experiments. It only discusses the experimental setup without hardware specifications. |
| Software Dependencies | No | The paper mentions various tools and models used (e.g., Open Face, Adam optimizer, VGG), but it does not provide specific version numbers for any of the software dependencies, libraries, or frameworks used to replicate the experiment. |
| Experiment Setup | Yes | The learning rate for the generator and discriminator are set to 2e 5 and 1e 5 respectively. We use Adam (Kingma and Ba 2014) as the optimizer. Spectral Normalization (Miyato et al. 2018) is utilized for each convolution layer in the generator. We set λGAN = 10, λc = 5 and λlocal = 5 in the loss function. The Gaussian kernel variance of heatmaps is 3. |