Learning Motion Refinement for Unsupervised Face Animation

Authors: Jiale Tao, Shuhang Gu, Wen Li, Lixin Duan

NeurIPS 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on widely used benchmarks demonstrate that our method achieves the best results among state-of-the-art baselines.
Researcher Affiliation Academia School of Computer Science and Engineering, UESTC1 Shenzhen Institute for Advanced Study, UESTC2
Pseudocode No The paper describes its method through text and equations but does not include structured pseudocode or an algorithm block.
Open Source Code Yes Codes will be available at https://github.com/Jiale Tao/MRFA/
Open Datasets Yes We conduct experiments on the widely used Voxceleb1 [18] dataset and the recently collected more challenged Celeb V-HQ dataset [44].
Dataset Splits No Voxceleb1 is a talking head dataset consisting of 20047 videos, among which 19522 are used for training and 525 are used for testing.
Hardware Specification Yes We train our method for 100 epochs on four NVIDIA A100 GPU cards or eight NVIDIA 3090 GPU cards.
Software Dependencies No The paper mentions optimizers (Adam) and networks (VGG-19, Unet) but does not provide specific version numbers for software dependencies or libraries.
Experiment Setup Yes We train our method for 100 epochs... The number of keyoints is set to 10... The patch radius r is set to 3 and the number of iterations is set to 6. The Adam optimizer [16] is adopted with β1 = 0.5 and β2 = 0.999, the initial learning rate is set as 2 * 10^-4 and dropped by a factor of 10 at the end of 60th and 90th epoch.