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 [1].

Reconstruction Target Matters in Masked Image Modeling for Cross-Domain Few-Shot Learning

Authors: Ran Ma, Yixiong Zou, Yuhua Li, Ruixuan Li

AAAI 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experiments on four CDFSL datasets demonstrate that our method achieves state-of-the-art performance.
Researcher Affiliation Academia School of Computer Science and Technology, Huazhong University of Science and Technology EMAIL
Pseudocode No The paper describes the methodology using text and diagrams (Figure 6 and 7) but does not include any explicit pseudocode or algorithm blocks.
Open Source Code No The paper does not contain any explicit statement about releasing source code or a link to a code repository.
Open Datasets Yes Following the BSCD benchmark (Guo et al. 2020), we use mini Imagenet (Vinyals et al. 2016) as the source dataset and four cross-domain datasets as target datasets: Crop Disease (Mohanty, Hughes, and Salath e 2016), Euro SAT(Helber et al. 2019), ISIC2018 (Codella et al. 2018) and Chest X (Wang et al. 2017).
Dataset Splits Yes Mini Image Net has 100 natural image categories with 600 images each, split into 64 training, 16 validation, and 20 test categories. ... existing research (Snell, Swersky, and Zemel 2017; Guo et al. 2020) employ a k-way n-shot paradigm to sample from DT , forming small datasets (episodes) containing k classes with n training samples each.
Hardware Specification No The paper does not explicitly describe the specific hardware (e.g., GPU models, CPU types) used for running its experiments.
Software Dependencies No The paper does not provide specific version numbers for ancillary software dependencies such as programming languages, libraries, or frameworks.
Experiment Setup Yes During base class training, we train the model with Adam W (Loshchilov and Hutter 2019) with a learning rate of 0.001 for the classifier, 1e-7 for the backbone, and 1e-6 for the decoder. For novel-class fine-tuning, we discard the decoder and fine-tune the backbone using the SGD optimizer with a momentum of 0.9.