This repository provides an implementation of DenseNet (Densely Connected Convolutional Networks) in Keras, supporting both standard and BC (Bottleneck-Compressed) variants. The architecture features dense connectivity where each layer receives feature maps from all preceding layers, and uses growth rate parameters to control feature map expansion. The implementation achieves state-of-the-art performance on CIFAR-10 and CIFAR-100 datasets, with DenseNet-BC-190-40 reaching 3.46% error on CIFAR-10.