Improving object detection with deep convolutional networks via Bayesian optimization and structured prediction

By Yuting Zhang, Kihyuk Sohn, Ruben Villegas, Gang Pan, Honglak Lee. 2015

221 citations recorded by the source corpus.

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 demonstrate 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.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

Citations in the corpus

Object Detection Networks on Convolutional Feature Maps (2015, methodology): “5 76.7 68.1 Table 8. Detection results for PASCAL VOC 2012 test set reported by the evaluation server. The entries of R-CNN use pre-trained VGG models. The entries of BabyLearning, NUS NIN c2000, and [33] are based on the R-CNN framework. Data augmentation is a standard method for improving ConvNet ”
Taking the Human Out of the Loop: A Review of Bayesian Optimization (2016, methodology): “We can denote these data as a set of tuples D n = { ( a i , y i ) } ni =1 , where a i indicates which of the K drugs was administered and y i is 1 if the patient was cured and 0 otherwise.”
Learning in high-dimensional multimedia data: the state of the art (2015, background): “Instead of reducing the high-dimensional multimedia data to fit traditional machine learning algorithms, some experts proposed effective learning schemas directly applied to these high-dimensional data to conduct advanced multimedia applications [62, 66, 101].”
DeepID-Net: Object Detection with Deformable Part Based Convolutional Neural Networks (2017, unknown): “Because of the power on learning feature representation, deep models have been widely used for object recognition, detection and other vision tasks [1], [6], [7], [19], [20], [21], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47],”
Deep learning for visual understanding (2017, unknown): “For instance, FGS [69] addresses the localization problem via two methods: 1) develop a ne-grained search algorithm to iteratively optimize the location; 2) train a CNN classi er with a structured SVM objective to balance between classi cation and localization.”
End-to-End Training of Object Class Detectors for Mean Average Precision (2016, methodology): “Modern techniques for object detection [1,2, ?”
Exploit All the Layers: Fast and Accurate CNN Object Detector with Scale Dependent Pooling and Cascaded Rejection Classifiers (2016, methodology): “Leveraging on the representational power of CNN, a number of methods are proposed to detect objects in natural images using CNN [13, 14, 16, 7, 42].”
RIFD-CNN: Rotation-Invariant and Fisher Discriminative Convolutional Neural Networks for Object Detection (2016, methodology): “Since the groundbreaking success of deep convolutional neural networks (CNN) [1] in image classification task [2] on the ImageNet large scale visual recognition challenge (ILSVRC) [3, 4], CNN-based object detection methods have recently attracted a great deal of research interest and have achieved s”
Unsupervised Discovery of Object Landmarks as Structural Representations (2018, background): “Modern neural networks can learn latent representations to effectively solve various vision problems, including image classification [26, 53, 56, 20], segmentation [32, 40, 21], object detection [17, 80, 49], human pose estimation [39], 3D reconstruction [13, 67, 14], and image generation [25, 18, 4”
A deep learning approach for the analysis of masses in mammograms with minimal user intervention (2017, methodology): “), and P (fN+t|dN+t,BN+t−1, θEI) follows a Gaussian distribution (Zhang et al. (2015)).”
Black-Box Tuning for Language-Model-as-a-Service (2022, methodology): “…2016), etc. Due to their ability of addressing complex optimization tasks, DFO algorithms have achieved many impressive applications in automatic machine learning (Snoek et al., 2012), reinforcement learning (Salimans et al., 2017; Hu et al., 2017), objective detection (Zhang et al., 2015b), etc.”
Deep learning for visual understanding: A review (2016, methodology): “Recent developments [164,177,178,183,185] are mainly focused on these two aspects.”