Texture Reformer: Towards Fast and Universal Interactive Texture Transfer

Authors: Zhizhong Wang, Lei Zhao, Haibo Chen, Ailin Li, Zhiwen Zuo, Wei Xing, Dongming Lu2624-2632

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

Reproducibility Variable Result LLM Response
Research Type Experimental The experimental results on a variety of application scenarios demonstrate the effectiveness and superiority of our framework. And compared with the state-of-the-art interactive texture transfer algorithms, it not only achieves higher quality results but, more remarkably, also is 2-5 orders of magnitude faster. We apply our framework to many challenging interactive texture transfer tasks, and demonstrate its effectiveness and superiority through extensive comparisons with the state-of-the-art (SOTA) algorithms. In Table 1, we compare the running time with the competitors.
Researcher Affiliation Academia College of Computer Science and Technology, Zhejiang University {endywon, cszhl, cshbchen, liailin, zzwcs, wxing, ldm}@zju.edu.cn
Pseudocode No The paper describes procedures using numbered steps within paragraphs (e.g., in Section 3.1 for SGTW), but it does not present these as formal pseudocode blocks or clearly labeled algorithm sections.
Open Source Code Yes Code is available at https://github.com/EndyWon/Texture-Reformer.
Open Datasets Yes The decoders are trained on the Microsoft COCO dataset (Lin et al. 2014)
Dataset Splits No The paper mentions training on the Microsoft COCO dataset but does not specify any dataset splits (e.g., percentages or counts for training, validation, or testing data) needed for reproduction.
Hardware Specification Yes 1 Tested on a 3.3 GHz hexa-core CPU and a 6GB Nvidia 1060 GPU.
Software Dependencies No The paper does not provide specific version numbers for software dependencies such as libraries, frameworks (e.g., PyTorch, TensorFlow), or programming languages beyond the general mentions in the text.
Experiment Setup Yes The hyperparameters that control the semantic-awareness (Eq. 4) in stage I and stage II are set to ω1 = ω2 = 50 (ω1 for stage I, ω2 for stage II. See supplementary material (SM) for their effects).