DFVSR: Directional Frequency Video Super-Resolution via Asymmetric and Enhancement Alignment Network

Authors: Shuting Dong, Feng Lu, Zhe Wu, Chun Yuan

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

Reproducibility Variable Result LLM Response
Research Type Experimental 4 Experiments, 4.1 Datasets and Implementation, 4.2 Comparison with State-of-The-Art Methods, 4.3 Ablation Study, Table 1: Quantitative comparison (PSNR and SSIM) of different methods on REDS4, Vimeo-T, Vid4 and UDM10 with upscale factor 4 under BI and BD degradations.
Researcher Affiliation Collaboration Shuting Dong1,2 , Feng Lu1,2 , Zhe Wu2 and Chun Yuan1,2 1Tsinghua Shenzhen International Graduate School, Tsinghua University 2Peng Cheng Laboratory
Pseudocode No The paper describes its proposed network and modules using textual descriptions and architectural diagrams (Figure 1 and Figure 2), but it does not include any formal pseudocode or algorithm blocks.
Open Source Code No The paper does not include any explicit statement about releasing source code for the methodology or a link to a code repository.
Open Datasets Yes We adopt two widely used datasets to train: REDS [Nah et al., 2019] and Vimeo-90K [Xue et al., 2019].
Dataset Splits Yes Following [Chan et al., 2021a], we apply REDS4 as our test set, and REDSval4 as the validation set.
Hardware Specification Yes The model is trained under the Py Torch framework with an NVIDIA RTX 2080Ti GPU.
Software Dependencies No The paper mentions training under the 'PyTorch framework' but does not provide specific version numbers for PyTorch or any other software dependencies.
Experiment Setup Yes We employ Adam optimizer by setting β1 = 0.9 and β2 = 0.999. The learning rate is initialized as 2.5 10 5. We apply RGB patches of size 64 64 as inputs. We set the mini-batch size to 32. In addition to our proposed DFLoss, we also adopt Charbonnier loss [Lai et al., 2017], and ε is set to 1 10 3. The total number of iterations is 600K.