ShuffleMixer: An Efficient ConvNet for Image Super-Resolution

Authors: Long Sun, Jinshan Pan, Jinhui Tang

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

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrate that the proposed Shuffle Mixer is about 3 smaller than the state-of-the-art efficient SR methods, e.g. CARN, in terms of model parameters and FLOPs while achieving competitive performance.
Researcher Affiliation Academia Long Sun, Jinshan Pan , Jinhui Tang Nanjing University of Science and Technology {cs.longsun, jspan, jinhuitang}@njust.edu.cn
Pseudocode No The paper describes the network architecture with diagrams but does not include any structured pseudocode or algorithm blocks.
Open Source Code Yes The code is available at https://github.com/sunny2109/Shuffle Mixer.
Open Datasets Yes Following existing methods [22, 24, 23], we train our models on the DF2K dataset, a merged dataset with DIV2K [37] and Flickr2K [25], which contains 3450 (800 + 2650) high-quality images.
Dataset Splits Yes Table 2: Ablation studies of the shuffler mixer layer and the feature mixing block on 4 DIV2K validation set[37].
Hardware Specification Yes All experiments are conducted with the Py Torch framework on an Nvidia Tesla V100 GPU.
Software Dependencies No The paper mentions 'PyTorch framework' but does not specify its version number or any other software dependencies with version details.
Experiment Setup Yes In each training mini-batch, we randomly crop 64 patches of size 64 64 from LR images as the input. The proposed model is trained by minimizing L1 loss and the frequency loss [5] with Adam [19] optimizer for 300,000 total iterations. The learning rate is set to a constant 5 10 4.