Implicit Warping for Animation with Image Sets

Authors: Arun Mallya, Ting-Chun Wang, Ming-Yu Liu

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We perform our ablations on the Talking Head-1KH [41] dataset at the 256ˆ256 resolution and compare with baselines at the full 512 ˆ 512 resolution. We also report results on the 256 ˆ 256 Vox Celeb2 [26] dataset. Additionally, to demonstrate the generality of our method, we report results on the more challenging TED Talk [28] dataset of moving upper bodies at a resolution of 384 ˆ 384. Metrics. We measure the fidelity of driving image reconstruction using PSNR, L1, LPIPS [53], and FID [12].
Researcher Affiliation Industry Arun Mallya amallya@nvidia.com Ting-Chun Wang tingchunw@nvidia.com Ming-Yu Liu mingyul@nvidia.com
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No We will release the code after acceptance.
Open Datasets Yes Datasets. We perform our ablations on the Talking Head-1KH [41] dataset at the 256ˆ256 resolution and compare with baselines at the full 512 ˆ 512 resolution. We also report results on the 256 ˆ 256 Vox Celeb2 [26] dataset. Additionally, to demonstrate the generality of our method, we report results on the more challenging TED Talk [28] dataset of moving upper bodies at a resolution of 384 ˆ 384.
Dataset Splits Yes Did you specify all the training details (e.g., data splits, hyperparameters, how they were chosen)? [Yes] See provided supplementary material.
Hardware Specification No The paper mentions performance in terms of FPS ('reasonably fast ( 10 FPS on 512 ˆ 512 images)') but does not specify the hardware (e.g., GPU model, CPU, memory) used for experiments.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers.
Experiment Setup Yes Did you specify all the training details (e.g., data splits, hyperparameters, how they were chosen)? [Yes] See provided supplementary material.