ProNet: Learning to Propose Object-specific Boxes for Cascaded Neural Networks

Abstract

This paper aims to classify and locate objects accurately and efficiently, without using bounding box annotations. It is challenging as objects in the wild could appear at arbitrary locations and in different scales. In this paper, we propose a novel classification architecture ProNet based on convolutional neural networks. It uses computationally efficient neural networks to propose image regions that are likely to contain objects, and applies more powerful but slower networks on the proposed regions. The basic building block is a multi-scale fully-convolutional network which assigns object confidence scores to boxes at different locations and scales. We show that such networks can be trained effectively using image-level annotations, and can be connected into cascades or trees for efficient object classification. ProNet outperforms previous state-of-the-art significantly on PASCAL VOC 2012 and MS COCO datasets for object classification and point-based localization.

CVPR 2016 (fixed reference issue)

References in corpus (7)

• Fully Convolutional Networks for Semantic Segmentation
• Return of the Devil in the Details: Delving Deep into Convolutional Nets
• Fully Convolutional Multi-Class Multiple Instance Learning
• Learning Deep Features for Discriminative Localization
• BoxSup: Exploiting Bounding Boxes to Supervise Convolutional Networks for Semantic Segmentation
• A Discriminative CNN Video Representation for Event Detection
• From Image-level to Pixel-level Labeling with Convolutional Networks

Cited by in corpus (4)

• Combining Fully Convolutional and Recurrent Neural Networks for 3D Biomedical Image Segmentation
• Deep Weakly-Supervised Learning Methods for Classification and Localization in Histology Images: A Survey
• Dynamic Deep Neural Networks: Optimizing Accuracy-Efficiency Trade-offs by Selective Execution
• Hard-Aware Deeply Cascaded Embedding