What question did this study set out to answer?

This research aims to enhance object detection performance by utilizing a deep CNN ensemble and data augmentation.

June 24, 2015Open Access

Deep CNN Ensemble with Data Augmentation for Object Detection

Key Points

This research aims to enhance object detection performance by utilizing a deep CNN ensemble and data augmentation.
Constructed an ensemble of deep CNN models with complementary architectures.
Augmented the PASCAL VOC training set with a selected subset of Microsoft COCO images.
Evaluated the method on the PASCAL VOC 2012 detection challenge.
Achieved state-of-the-art performance on the PASCAL VOC 2012 detection task.
Outperformed all previous methods submitted at the time.
Significant training data enlargement led to improved accuracy.

Abstract

We report on the methods used in our recent DeepEnsembleCoco submission to the PASCAL VOC 2012 challenge, which achieves state-of-the-art performance on the object detection task. Our method is a variant of the R-CNN model proposed Girshick:CVPR14 with two key improvements to training and evaluation. First, our method constructs an ensemble of deep CNN models with different architectures that are complementary to each other. Second, we augment the PASCAL VOC training set with images from the Microsoft COCO dataset to significantly enlarge the amount training data. Importantly, we select a subset of the Microsoft COCO images to be consistent with the PASCAL VOC task. Results on the PASCAL VOC evaluation server show that our proposed method outperform all previous methods on the PASCAL VOC 2012 detection task at time of submission.

اسأل الذكاء الاصطناعي

Bookmark

View Full Paper