Scale-Invariant Object Detection by Adaptive Convolution with Unified Global-Local Context

TL;DR

This paper introduces SAC-Net, a novel adaptive convolutional network that dynamically adjusts atrous rates and incorporates global context, significantly improving multi-scale object detection, especially for small objects, over existing models.

Contribution

The paper proposes a switchable, adaptive atrous convolution mechanism within a unified framework, enhancing dense feature retention and scale-invariance in object detection.

Findings

Outperforms state-of-the-art models on benchmark datasets

Improves detection accuracy for small and multi-scale objects

Demonstrates robustness with dynamic atrous rate adjustment

Abstract

Dense features are important for detecting minute objects in images. Unfortunately, despite the remarkable efficacy of the CNN models in multi-scale object detection, CNN models often fail to detect smaller objects in images due to the loss of dense features during the pooling process. Atrous convolution addresses this issue by applying sparse kernels. However, sparse kernels often can lose the multi-scale detection efficacy of the CNN model. In this paper, we propose an object detection model using a Switchable (adaptive) Atrous Convolutional Network (SAC-Net) based on the efficientDet model. A fixed atrous rate limits the performance of the CNN models in the convolutional layers. To overcome this limitation, we introduce a switchable mechanism that allows for dynamically adjusting the atrous rate during the forward pass. The proposed SAC-Net encapsulates the benefits of both low-level and high-level features to achieve improved performance on multi-scale object detection tasks, without losing the dense features. Further, we apply a depth-wise switchable atrous rate to the proposed network, to improve the scale-invariant features. Finally, we apply global context on the proposed model. Our extensive experiments on benchmark datasets demonstrate that the proposed SAC-Net outperforms the state-of-the-art models by a significant margin in terms of accuracy.