Reproduce “Depth-guided Geometric Adaptation for Few-shot Image Classification” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zhiqi Miao, Estefanía Talavera, Weichuan Zhang, Jiapan Guo
DAFIC addresses few-shot image classification by using estimated depth as a soft cue to guide patch relevance and bidirectional patch matching. In MiniImageNet 5-way 1-shot episodes, it reports mean accuracy of 77.71% ± 0.44%, compared with 74.28% ± 0.18% for SMKD with the same ViT-S backbone. This result is on the reported benchmark, not a general guarantee.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.