Feature Selective Networks for Object Detection

TL;DR

This paper introduces feature selective networks that enhance object detection by exploiting sub-region and aspect ratio disparities in RoI features, leading to significant performance improvements across multiple datasets.

Contribution

The proposed network reformulates RoI features using sub-region and aspect ratio attention banks, improving detection accuracy with minimal additional complexity.

Findings

Over 3% mAP improvement on PASCAL VOC and MS COCO datasets.

Effective enhancement across various backbone networks like ResNet-101, GoogLeNet, and VGG-16.

Simple integration yields consistent detection performance boosts.

Abstract

Objects for detection usually have distinct characteristics in different sub-regions and different aspect ratios. However, in prevalent two-stage object detection methods, Region-of-Interest (RoI) features are extracted by RoI pooling with little emphasis on these translation-variant feature components. We present feature selective networks to reform the feature representations of RoIs by exploiting their disparities among sub-regions and aspect ratios. Our network produces the sub-region attention bank and aspect ratio attention bank for the whole image. The RoI-based sub-region attention map and aspect ratio attention map are selectively pooled from the banks, and then used to refine the original RoI features for RoI classification. Equipped with a light-weight detection subnetwork, our network gets a consistent boost in detection performance based on general ConvNet backbones (ResNet-101, GoogLeNet and VGG-16). Without bells and whistles, our detectors equipped with ResNet-101 achieve more than 3% mAP improvement compared to counterparts on PASCAL VOC 2007, PASCAL VOC 2012 and MS COCO datasets.