Improving Object Detection with Deep Convolutional Networks via Bayesian Optimization and Structured Prediction

Abstract

Object detection systems based on the deep convolutional neural network (CNN) have recently made ground- breaking advances on several object detection benchmarks. While the features learned by these high-capacity neural networks are discriminative for categorization, inaccurate localization is still a major source of error for detection. Building upon high-capacity CNN architectures, we address the localization problem by 1) using a search algorithm based on Bayesian optimization that sequentially proposes candidate regions for an object bounding box, and 2) training the CNN with a structured loss that explicitly penalizes the localization inaccuracy. In experiments, we demonstrated that each of the proposed methods improves the detection performance over the baseline method on PASCAL VOC 2007 and 2012 datasets. Furthermore, two methods are complementary and significantly outperform the previous state-of-the-art when combined.

CVPR 2015

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Deep Learning in Neural Networks: An Overview
• Practical Bayesian Optimization of Machine Learning Algorithms
• Caffe: Convolutional Architecture for Fast Feature Embedding
• Going Deeper with Convolutions
• Improving Object Detection with Deep Convolutional Networks via Bayesian Optimization and Structured Prediction

Cited by in corpus (28)

• Improving Object Detection with Deep Convolutional Networks via Bayesian Optimization and Structured Prediction
• Recent Advances in Object Detection in the Age of Deep Convolutional Neural Networks
• Object Detection with Deep Learning: A Review
• Bayesian Optimization of Combinatorial Structures
• Exploiting Depth from Single Monocular Images for Object Detection and Semantic Segmentation
• RON: Reverse Connection with Objectness Prior Networks for Object Detection
• Derivative-Free Reinforcement Learning: A Review
• Object Detection Networks on Convolutional Feature Maps
• A Survey of Deep Learning Techniques for Mobile Robot Applications
• Exploring Person Context and Local Scene Context for Object Detection
• Co-localization with Category-Consistent Features and Geodesic Distance Propagation
• Road Crack Detection Using Deep Convolutional Neural Network and Adaptive Thresholding
• Dynamic Zoom-in Network for Fast Object Detection in Large Images
• End-to-end training of object class detectors for mean average precision
• Multi-Object Classification and Unsupervised Scene Understanding Using Deep Learning Features and Latent Tree Probabilistic Models
• A Survey on Deep Domain Adaptation and Tiny Object Detection Challenges, Techniques and Datasets
• Learning in High-Dimensional Multimedia Data: The State of the Art
• Leveraging End-to-End Speech Recognition with Neural Architecture Search
• Unsupervised Discovery of Object Landmarks as Structural Representations
• Factors in Finetuning Deep Model for object detection
• Action-Driven Object Detection with Top-Down Visual Attentions
• Reinforced Few-Shot Acquisition Function Learning for Bayesian Optimization
• Cost-Sensitive Deep Learning with Layer-Wise Cost Estimation
• Discriminative Bimodal Networks for Visual Localization and Detection with Natural Language Queries
• Adaptively Denoising Proposal Collection for Weakly Supervised Object Localization
• RoomStructNet: Learning to Rank Non-Cuboidal Room Layouts From Single View
• cvpaper.challenge in 2016: Futuristic Computer Vision through 1,600 Papers Survey
• cvpaper.challenge in 2015 - A review of CVPR2015 and DeepSurvey