• Corpus ID: 8827762

Deep Neural Networks for Object Detection

@inproceedings{Szegedy2013DeepNN,
  title={Deep Neural Networks for Object Detection},
  author={Christian Szegedy and Alexander Toshev and D. Erhan},
  booktitle={NIPS},
  year={2013}
}
Deep Neural Networks (DNNs) have recently shown outstanding performance on image classification tasks [14. [] Key Method We define a multi-scale inference procedure which is able to produce high-resolution object detections at a low cost by a few network applications. State-of-the-art performance of the approach is shown on Pascal VOC.
View Paper
machinelearning.wustl.edu
1,203 Citations
Highly Influential Citations
44
Background Citations
495
Methods Citations
276
Results Citations
6

Figures and Tables from this paper

1,203 Citations

Structured Prediction for Object Detection in Deep Neural Networks
TLDR
This work proposes to adapt a structured loss function for neural network training which directly maximizes overlap of the prediction with ground truth bounding boxes, and shows how this structured loss can be implemented efficiently.
Deep learning for class-generic object detection
TLDR
It is shown that neural networks originally designed for image recognition can be trained to detect objects within images, regardless of their class, including objects for which no bounding box labels have been provided.
Scalable Object Detection Using Deep Neural Networks
TLDR
This work proposes a saliency-inspired neural network model for detection, which predicts a set of class-agnostic bounding boxes along with a single score for each box, corresponding to its likelihood of containing any object of interest.
A Novel Representation and Pipeline for Object Detection
TLDR
A novel training criterion is proposed which tackles background separately for object detection and how Learning without Forgetting and finetuning perform in transferring from the classification to the detection task is examined.
Improving object detection with deep convolutional networks via Bayesian optimization and structured prediction
TLDR
This work addresses the localization problem by using a search algorithm based on Bayesian optimization that sequentially proposes candidate regions for an object bounding box, and training the CNN with a structured loss that explicitly penalizes the localization inaccuracy.
AttentionNet: Aggregating Weak Directions for Accurate Object Detection
TLDR
AttentionNet is presented, a novel detection method using a deep convolutional neural network, named AttentionNet, which detects objects without any separated models from the object proposal to the post bounding-box regression.
Learning to decompose for object detection and instance segmentation
TLDR
This work proposes a novel end-to-end trainable deep neural network architecture that generates the correct number of object instances and their bounding boxes (or segmentation masks) given an image, using only a single network evaluation without any pre- or post-processing steps.
Efficient ConvNet for Surface Object Recognition
TLDR
A dynamic-selecting criterion approach is proposed to prune a trained Yolo-v2 model to deal with drawbacks caused by redundant parameters in network and it is shown that this approach can reduce inference costs for Yolo -v2 by up to 65% on it while regaining close to the original performance by retraining the network.
Improving object detection via improving accuracy of object localization
TLDR
This work combines a high-recall algorithm proposing candidate regions for an object bounding box with an algorithm reducing localization bias, and utilizing box alignment which penalizing deviation via taking object boundaries into account, to instruct the procedure of proposing input of CNN.
Part Detector Discovery in Deep Convolutional Neural Networks
TLDR
This paper shows how pre-trained convolutional neural networks can be used for robust and efficient object part discovery and localization without the necessity to actually train the network on the current dataset.
...
...

References

SHOWING 1-10 OF 21 REFERENCES
Object Detection with Discriminatively Trained Part Based Models
We describe an object detection system based on mixtures of multiscale deformable part models. Our system is able to represent highly variable object classes and achieves state-of-the-art results in…
The Pascal Visual Object Classes (VOC) Challenge
TLDR
The state-of-the-art in evaluated methods for both classification and detection are reviewed, whether the methods are statistically different, what they are learning from the images, and what the methods find easy or confuse.
Inference and Learning with Hierarchical Shape Models
TLDR
This work automates the decomposition of an object category into parts and contours, and discriminatively learn the cost function that drives the matching of the object to the image using Multiple Instance Learning.
Building high-level features using large scale unsupervised learning
TLDR
Contrary to what appears to be a widely-held intuition, the experimental results reveal that it is possible to train a face detector without having to label images as containing a face or not.
Learning Hierarchical Features for Scene Labeling
TLDR
A method that uses a multiscale convolutional network trained from raw pixels to extract dense feature vectors that encode regions of multiple sizes centered on each pixel, alleviates the need for engineered features, and produces a powerful representation that captures texture, shape, and contextual information.
Latent hierarchical structural learning for object detection
TLDR
This paper describes an incremental concave-convex procedure (iCCCP) which allows us to learn both two and three layer models efficiently and demonstrates the advantages of three layer hierarchies - outperforming Felzenszwalb et al.'s two layer models on all 20 classes.
Histograms of oriented gradients for human detection
  • N. Dalal, B. Triggs
  • Computer Science
    2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)
  • 2005
TLDR
It is shown experimentally that grids of histograms of oriented gradient (HOG) descriptors significantly outperform existing feature sets for human detection, and the influence of each stage of the computation on performance is studied.
Learning Deep Architectures for AI
TLDR
The motivations and principles regarding learning algorithms for deep architectures, in particular those exploiting as building blocks unsupervised learning of single-layer modelssuch as Restricted Boltzmann Machines, used to construct deeper models such as Deep Belief Networks are discussed.
Deep Neural Networks Segment Neuronal Membranes in Electron Microscopy Images
TLDR
This work addresses a central problem of neuroanatomy, namely, the automatic segmentation of neuronal structures depicted in stacks of electron microscopy images, using a special type of deep artificial neural network as a pixel classifier to segment biological neuron membranes.
Learning the Taxonomy and Models of Categories Present in Arbitrary Images
  • N. Ahuja, S. Todorovic
  • Computer Science
    2007 IEEE 11th International Conference on Computer Vision
  • 2007
TLDR
This paper proposes, and presents a solution to, the problem of simultaneous learning of multiple visual categories present in an arbitrary image set and their inter-category relationships, and proposes a directed acyclic graph (DAG) which represents the discovered taxonomy.
...
...