CondConv: Conditionally Parameterized Convolutions for Efficient Inference

TL;DR

CondConv introduces conditionally parameterized convolutions that learn specialized kernels per example, enhancing neural network capacity and performance while maintaining efficient inference, demonstrated on ImageNet classification.

Contribution

This paper presents CondConv, a novel convolutional layer that learns per-example kernels, enabling larger, more capable networks without increased inference cost.

Findings

Improved accuracy on ImageNet with EfficientNet-B0 using CondConv.

Achieved 78.3% top-1 accuracy with only 413M multiply-adds.

Enhanced performance on detection tasks with CondConv.

Abstract

Convolutional layers are one of the basic building blocks of modern deep neural networks. One fundamental assumption is that convolutional kernels should be shared for all examples in a dataset. We propose conditionally parameterized convolutions (CondConv), which learn specialized convolutional kernels for each example. Replacing normal convolutions with CondConv enables us to increase the size and capacity of a network, while maintaining efficient inference. We demonstrate that scaling networks with CondConv improves the performance and inference cost trade-off of several existing convolutional neural network architectures on both classification and detection tasks. On ImageNet classification, our CondConv approach applied to EfficientNet-B0 achieves state-of-the-art performance of 78.3% accuracy with only 413M multiply-adds. Code and checkpoints for the CondConv Tensorflow layer and CondConv-EfficientNet models are available at: https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet/condconv.