Reproduce “Modified Deep Neural Networks for Dog Breeds Identification” 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
Aydin Ayanzadeh, Sahand Vahidnia
The authors evaluate transfer learning for dog-breed classification by adapting ImageNet-pretrained models and augmenting training images. On Stanford Dogs, with 120 breeds and 20,580 images, ResNet-50 reaches 89.66% test accuracy after fine-tuning plus augmentation, versus 73.28% for its unmodified baseline. This bundled comparison does not isolate either component or establish superiority over prior systems.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.