Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
DENet: Disentangled Embedding Network for Visible Watermark Removal
Authors: Ruizhou Sun, Yukun Su, Qingyao Wu
AAAI 2023 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments are conducted on different challenging benchmarks. Experimental evaluations show that our approach can achieve state-of-the-art performance and yield high-quality images. |
| Researcher Affiliation | Academia | 1School of Software Engineering, South China University of Technology 2Key Laboratory of Big Data and Intelligent Robot, Ministry of Education 3Pazhou Lab, Guangzhou, China 4Peng Cheng Laboratory, China |
| Pseudocode | No | No pseudocode or clearly labeled algorithm blocks were found in the paper. |
| Open Source Code | Yes | The code is available at: https://github.com/lianchengmingjue/DENet. |
| Open Datasets | Yes | Similar to the existing watermark removal method (Cun and Pun 2021), all the experiments are conducted on the LOGO series dataset. |
| Dataset Splits | No | The paper specifies training (12151 images) and testing (2025 images) splits for the LOGO series datasets, but does not explicitly mention a separate validation dataset split. |
| Hardware Specification | No | No specific hardware details (e.g., GPU/CPU models, memory amounts, or detailed computer specifications) used for running experiments were provided in the paper. |
| Software Dependencies | No | The paper states 'Our method is implemented with Pytorch (Paszke et al. 2019)' but does not provide a specific version number for Pytorch or any other software dependencies. |
| Experiment Setup | Yes | The models are trained for 200 epochs, where the input image resolution is 256 256. We choose Adam (Kingma and Ba 2014) as optimizer with learning rate of 1e-3, batch size 16. The hyper-parameters in (7) are λmask = 1, λvgg = 0.25, λcontrast = 0.25, respectively. Therefore, we set temperature parameter τ to 0.07 and multi-head numbers h to 4 in the following experiments. |