Greatness in Simplicity: Unified Self-Cycle Consistency for Parser-Free Virtual Try-On

Authors: Chenghu Du, junyin Wang, Shuqing Liu, Shengwu Xiong

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments demonstrate that our method achieves state-of-the-art performance on a popular virtual try-on benchmark. We conduct experiments using the VITON dataset [8]
Researcher Affiliation Academia 1Wuhan University of Technology, 2Shanghai AI Laboratory 3Sanya Science and Education Innovation Park, Wuhan University of Technology 4Wuhan Textile University, 5Qiongtai Normal University
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper provides a project page link (https://du-chenghu.github.io/USC-PFN/) but does not explicitly state that the source code is provided there or through any other means within the paper itself.
Open Datasets Yes We conduct experiments using the VITON dataset [8]
Dataset Splits No The paper mentions 'training set' and 'test set' but does not explicitly state details for a validation set or its split.
Hardware Specification Yes The USC-PFN is implemented in Py Torch and trained on a single Nvidia Tesla V100 GPU running Ubuntu 16.04.
Software Dependencies No The paper mentions 'Py Torch' and 'Ubuntu 16.04' but does not provide specific version numbers for these or other ancillary software dependencies.
Experiment Setup Yes During training, a batch size of 16 is used for 100 epochs, and the Adam optimizer [29] is employed with parameters β1 = 0.5 and β2 = 0.999, and the initial learning rate is set to 1e 4 with linear decay after 50 epochs. ... In the loss functions, the λr = 20 and λp = 0.25 in the Lngd. The λscyc = 1, λG adv = 0.1, λsr = 50, λgr = 1, and λcp = 10 in the Lt sig.