Machine Learning Application Benchmark - DenseNet121

This is a re-implementation of DenseNet121 as [1] changed to fulfill hardware constraints of the Vitis AI framework for inference on Xilinx FPGAs.

Pretrained weights are available from the model database as follows:

• imagenet_densenet121.h5: /weights/imagenet_densenet121.h5
• imagenet_densenet121_notop.h5: /weights/imagenet_densenet121_notop.h5

Source file: ../amodels/densenet121_vitis.py

Usage:

models.densenet121_vitis.densenet121_vitis(
    input_tensor=None, 
    include_top=True, 
    weight_path=None, 
    return_tensor=False, 
    classes=1000, 
    classifier_activation="softmax",
    grow_rate=32, 
    compression=0.5, 
    kernel_size_first_layer=(7,7)
)

Arguments:

• input_tensor: optional keras layer, like an input tensor.
• include_top: whether to include the top layers or top.
• weight_path: If not none, these weights will be loaded.
• return_tensor: Whether to return the network as tensor or as tf.keras.model (if true, weights will not be loaded).
• classes: By default the number of classes are 1000 (ImageNet). Only important include_top=True.
• classifier_activation: By default softmax (ImageNet). Only important if include_top=True.
• grow_rate: Graw rate of the network, by default 32.
• compression: Compression factor of the transition layer blocks, by default 0.5
• kernel_size_first_layer: Kernel size of the first layer, by default (7,7)

Returns:

The CNN architecture as tf.keras.model if return_tensor=False, otherwise as tf.keras.layers.

References:

[1] G. Huang, Z. Liu, L. Van Der Maaten and K. Q. Weinberger, “Densely Connected Convolutional Networks,” 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 2261-2269, doi: 10.1109/CVPR.2017.243.
[2] O. Russakovsky et al., “ImageNet Large Scale Visual Recognition Challenge,” International Journal of Computer Vision (IJCV), vol. 115, no. 3, pp. 211–252, 2015, doi: 10.1007/s11263-015-0816-y.