Learning Chained Deep Features and Classifiers for Cascade in Object Detection

TL;DR

This paper introduces CC-Net, a chained cascade network that leverages previous stage features and classifiers to improve object detection accuracy and speed, achieving state-of-the-art results on PASCAL VOC 2007.

Contribution

The paper proposes a novel chained cascade framework with feature chaining and joint end-to-end training for improved object detection.

Findings

Achieved 81.1% mAP on PASCAL VOC 2007

Boosted detection performance on benchmarks

Demonstrated effectiveness of feature and classifier chaining

Abstract

Cascade is a widely used approach that rejects obvious negative samples at early stages for learning better classifier and faster inference. This paper presents chained cascade network (CC-Net). In this CC-Net, the cascaded classifier at a stage is aided by the classification scores in previous stages. Feature chaining is further proposed so that the feature learning for the current cascade stage uses the features in previous stages as the prior information. The chained ConvNet features and classifiers of multiple stages are jointly learned in an end-to-end network. In this way, features and classifiers at latter stages handle more difficult samples with the help of features and classifiers in previous stages. It yields consistent boost in detection performance on benchmarks like PASCAL VOC 2007 and ImageNet. Combined with better region proposal, CC-Net leads to state-of-the-art result of 81.1% mAP on PASCAL VOC 2007.