CPWC: Contextual Point Wise Convolution for Object Recognition

TL;DR

This paper introduces CPWC, a novel pointwise convolution that incorporates spatial context, significantly enhancing network performance in classification and detection tasks without greatly increasing parameters.

Contribution

The paper proposes a new spatially-aware pointwise convolution (CPWC) that improves deep network performance by utilizing surrounding spatial information efficiently.

Findings

Improved classification accuracy with CPWC.

Enhanced detection performance using the proposed method.

Significant performance gains without large parameter increase.

Abstract

Convolutional layers are a major driving force behind the successes of deep learning. Pointwise convolution (PWC) is a 1x1 convolutional filter that is primarily used for parameter reduction. However, the PWC ignores the spatial information around the points it is processing. This design is by choice, in order to reduce the overall parameters and computations. However, we hypothesize that this shortcoming of PWC has a significant impact on the network performance. We propose an alternative design for pointwise convolution, which uses spatial information from the input efficiently. Our design significantly improves the performance of the networks without substantially increasing the number of parameters and computations. We experimentally show that our design results in significant improvement in the performance of the network for classification as well as detection.