Learning with Recursive Perceptual Representations
By Oriol Vinyals, Yangqing Jia, Li Deng, Trevor Darrell. 2012
69 citations recorded by the source corpus.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
Citations in the corpus
- Towards Deep Neural Network Architectures Robust to Adversarial Examples (2014, methodology): “In practice, SVMs can be used to replace the final soft-max layer in classifier neural networks leading to better generalization Tang (2013), but applying SVM in the manifold space does not guarantee local generalization in the input space.”
- Deep Learning: Methods and Applications (2014, background): “In more recent work of [106, 110, 218, 366, 377], a new deep learning architecture, sometimes called deep stacking network (DSN), together with its tensor variant [180, 181] and its kernel version [102], are developed that all focus on discrimination with scalable, parallelizable, block-wise learnin”
- Deep Learning using Linear Support Vector Machines (2013, background): “Vinyals et al. (2012) learns a recursive representation using linear SVMs at every layer, but without joint fine-tuning of the hidden representation.”
- Embedding Entities and Relations for Learning and Inference in Knowledge Bases (2014, background): “As learning representations using deep networks has shown great success in various applications (Hinton et al., 2012; Vinyals et al., 2012; Deng et al., 2013), it may also help capturing hierarchical structure hidden in the multi-relational data.”
- A tutorial survey of architectures, algorithms, and applications for deep learning (2014, background): “(Notable exceptions are the recent kernel meth-ods that have been inspired by and integrated with deep learning; e.g., [10–12].)”
- Particle swarm optimization for hyper-parameter selection in deep neural networks (2017, unknown): “14 Recursive perceptual representations [34] 20.”
- Machine Learning Paradigms for Speech Recognition: An Overview (2013, unknown): “, [20], [108], [239]–[256].”
- Deep Learning using Support Vector Machines (2013, background): “Vinyals et al. (2012) learns a recursive representation using linear SVMs at every layer, but without joint fine-tuning of the hidden represen- tation.”
- Sequence classification using the high-level features extracted from deep neural networks (2014, methodology): “The method of building deep stacking networks is a more general, recursive way of creating tandem-like features as the overall network built gets deeper and deeper [10][11][20][33].”
- A Comprehensive Study on Center Loss for Deep Face Recognition (2019, background): “…loss is widely used in close-set recognition tasks, like object (Krizhevsky et al. 2012), scene (Zhou et al. 2014a), action (Wang et al. 2015b), etc. Alternatively, Vinyals et al. (2012), Nagi et al. (2012), Tang (2013) applied hinge loss in CNNs, which provided more geometric interpretation.”