Corpus ID: 6161478

DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition

@inproceedings{Donahue2014DeCAFAD,
  title={DeCAF: A Deep Convolutional Activation Feature for Generic Visual Recognition},
  author={Jeff Donahue and Yangqing Jia and Oriol Vinyals and Judy Hoffman and Ning Zhang and Eric Tzeng and Trevor Darrell},
  booktitle={ICML},
  year={2014}
}
We evaluate whether features extracted from the activation of a deep convolutional network trained in a fully supervised fashion on a large, fixed set of object recognition tasks can be repurposed to novel generic tasks. Our generic tasks may differ significantly from the originally trained tasks and there may be insufficient labeled or unlabeled data to conventionally train or adapt a deep architecture to the new tasks. We investigate and visualize the semantic clustering of deep convolutional… Expand
Deep convolutional neural networks as generic feature extractors
TLDR
These findings indicate that convolutional networks are able to learn generic feature extractors that can be used for different tasks, and also indicate that the long time needed to train such deep networks is a major drawback. Expand
From generic to specific deep representations for visual recognition
TLDR
This paper thoroughly investigates the transferability of ConvNet representations w.r.t. several factors, and shows that different visual recognition tasks can be categorically ordered based on their distance from the source task. Expand
Factors of Transferability for a Generic ConvNet Representation
TLDR
This paper introduces and investigates several factors affecting the transferability of Convolutional Networks, and shows that significant improvements can be achieved on various visual recognition tasks. Expand
Multifaceted Analysis of Fine-Tuning in a Deep Model for Visual Recognition
TLDR
This article introduces and systematically investigates several factors that influence the performance of fine-tuning for visual recognition, and quantitatively and qualitatively analyze these factors, evaluating their influence, and presenting many empirical observations. Expand
Learning and Transferring Deep ConvNet Representations with Group-Sparse Factorization
Deep convolutional neural networks (Deep ConvNets or CNNs) have exhibited their promise as a universal image representation for recognition. In this work we explore how the transferability of suchExpand
Understanding and Designing Convolutional Networks for Local Recognition Problems
TLDR
This work presents evidence that convnet features localize at a much finer scale than their receptive field sizes, that they can be used to perform intraclass alignment as well as conventional hand-engineered features, and that they outperform conventional features in keypoint prediction on objects from PASCAL VOC 2011. Expand
Semantic transfer with deep neural networks
TLDR
This work considers two novel scenarios of visual learning where knowledge transfer is affected from off-the-shelf convolutional neural networks (CNNs), and proposes a holistic scene representation derived with the help of pre-trained object recognition neural nets. Expand
Deep Networks Can Resemble Human Feed-forward Vision in Invariant Object Recognition
TLDR
This work benchmarked eight state-of-the-art DCNNs, the HMAX model, and a baseline shallow model and compared their results to those of humans with backward masking to demonstrate that shallow nets can outperform deep nets and humans when variations are weak. Expand
Hierarchically Robust Representation Learning
  • Qi Qian, Juhua Hu, Hao Li
  • Computer Science
  • 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2020
TLDR
This work investigates the phenomenon and demonstrates that deep features can be suboptimal due to the fact that they are learned by minimizing the empirical risk, and proposes a hierarchically robust optimization method to learn more generic features. Expand
Deep Convolutional Features for Image Based Retrieval and Scene Categorization
TLDR
This paper examines several pooling strategies derived for CNN features and demonstrates superior performance on the image retrieval task (INRIA Holidays) at the fraction of the computational cost, while using a relatively small memory requirements. Expand
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 55 REFERENCES
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. Expand
Learning hierarchical invariant spatio-temporal features for action recognition with independent subspace analysis
TLDR
This paper presents an extension of the Independent Subspace Analysis algorithm to learn invariant spatio-temporal features from unlabeled video data and discovered that this method performs surprisingly well when combined with deep learning techniques such as stacking and convolution to learn hierarchical representations. Expand
Unsupervised and Transfer Learning Challenge: a Deep Learning Approach
TLDR
This paper describes different kinds of layers the authors trained for learning representations in the setting of the Unsupervised and Transfer Learning Challenge, and the particular one-layer learning algorithms feeding a simple linear classifier with a tiny number of labeled training samples. Expand
DLID: Deep Learning for Domain Adaptation by Interpolating between Domains
TLDR
A novel deep learning model for domain adaptation is proposed which attempts to learn a predictively useful representation of the data by taking into account information from the distribution shift between the training and test data. Expand
ImageNet classification with deep convolutional neural networks
TLDR
A large, deep convolutional neural network was trained to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 different classes and employed a recently developed regularization method called "dropout" that proved to be very effective. Expand
Adapting Visual Category Models to New Domains
TLDR
This paper introduces a method that adapts object models acquired in a particular visual domain to new imaging conditions by learning a transformation that minimizes the effect of domain-induced changes in the feature distribution. Expand
Transfer learning for image classification with sparse prototype representations
TLDR
The results show that when only few examples are available for training a target topic, leveraging knowledge learnt from other topics can significantly improve performance. Expand
Geodesic flow kernel for unsupervised domain adaptation
TLDR
This paper proposes a new kernel-based method that takes advantage of low-dimensional structures that are intrinsic to many vision datasets, and introduces a metric that reliably measures the adaptability between a pair of source and target domains. Expand
Towards Scalable Representations of Object Categories: Learning a Hierarchy of Parts
  • S. Fidler, A. Leonardis
  • Mathematics, Computer Science
  • 2007 IEEE Conference on Computer Vision and Pattern Recognition
  • 2007
TLDR
This paper proposes a novel approach to constructing a hierarchical representation of visual input that aims to enable recognition and detection of a large number of object categories by learning a hierarchy of spatially flexible compositions in an unsupervised, statistics-driven manner. Expand
Deformable Part Descriptors for Fine-Grained Recognition and Attribute Prediction
TLDR
This paper proposes two pose-normalized descriptors based on computationally-efficient deformable part models based on strongly-supervised DPM parts, which enable pooling across pose and viewpoint, in turn facilitating tasks such as fine-grained recognition and attribute prediction. Expand
...
1
2
3
4
5
...