Deep Learning of Invariant Features via Simulated Fixations in Video

By Will Y. Zou, Andrew Y. Ng, Shenghuo Zhu, Kai Yu. 2012

173 citations recorded by the source corpus.

Subfields: 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 Computer Vision and Pattern Recognition

Citations in the corpus

Discriminative Unsupervised Feature Learning with Convolutional Neural Networks (2014, methodology): “6 1024× 64 Slowness on videos [11] 61.”
Invariant Information Clustering for Unsupervised Image Classification and Segmentation (2018, background): “…or non-material distortion is an idea shared by IIC with several works, including exemplars [17], IMSAT [27], proximity prediction [29], the denoising objective of Tagger [21], temporal slowness constraints [53], and optimising for features to be invariant to local image transformations [45, 28].”
Multi-task Self-Supervised Visual Learning (2017, background): “In the last few years, a considerable number of such tasks have been proposed [1, 2, 6, 7, 8, 17, 20, 21, 23, 25, 26, 27, 28, 29, 31, 39, 40, 42, 43, 46, 47], such as asking a neural network to colorize grayscale images, fill in image holes, solve jigsaw puzzles made from image patches, or predict m”
Learning Image Representations Tied to Ego-Motion (2015, background): “Temporal coherence has been explored for unsupervised feature learning with CNNs [22, 37, 9, 3, 19], with applications to dimensionality reduction [10], object recognition [22, 37], and metric learning [9].”
Unsupervised Learning of Visual Representations Using Videos (2015, background): “Early work such as [56] focused on inclusion of constraints via video to autoencoder framework.”
Generative Image Modeling Using Style and Structure Adversarial Networks (2016, background): “There are two common approaches for unsupervised learning: (a) using a discriminative framework with auxiliary tasks where supervision comes for free, such as context prediction [1,2] or temporal embedding [3,4,5,6,7,8]; (b) using a generative framework where the underlying model is compositional an”
Unsupervised Learning of Long-Term Motion Dynamics for Videos (2017, background): “The most common constraint is enforcing learned representations to be temporally smooth [47].”
Stochastic Video Generation with a Learned Prior (2018, background): “Zou et al. (2012) use similar principles to learn features that exhibit a range of complex invariances.”
Unsupervised Learning of Spatiotemporally Coherent Metrics (2014, background): “Several other studies combine the slowness prior with independence inducing priors [17, 5, 22].”
Slow and Steady Feature Analysis: Higher Order Temporal Coherence in Video (2015, methodology): “The existing methods either produce a holistic image embedding [31, 3, 13, 15], or else track local patches to learn a localized representation [47, 48, 42].”
Temporal Generative Adversarial Nets with Singular Value Clipping (2016, methodology): “Deep neural networks have also been applied to feature extraction from videos [51, 6, 48] in the same way as supervised learning.”
Object-Centric Representation Learning from Unlabeled Videos (2016, methodology): “Some methods attempt to learn feature representations of video frames as a whole [6,20,9,10], while others track local patches to learn a localized representation [7,21,22].”
Deep Representation Learning with Target Coding (2015, unknown): “30 – (Zou et al. 2012):Simulated Fixations – – – – 61.”
Direct Modeling of Complex Invariances for Visual Object Features (2013, background): “Indeed, the current direction tends to require considerable additional data, such as video datasets (Zou et al., 2012), or in some cases massive amount of unlabeled data (Le et al., 2012; Coates et al., 2012).”
Towards Nonlinear Disentanglement in Natural Data with Temporal Sparse Coding (2020, methodology): “[46] Will Zou, Shenghuo Zhu, Kai Yu, and Andrew Y Ng.”
Hierarchical Convolutional Features for Visual Tracking (2015, background): “[29] propose to learn a two-layer neural network on a video repository [39], where temporally slowness constraints are imposed for feature learning.”
Unsupervised Learning of Discriminative Attributes and Visual Representations (2016, background): “Many unsupervised studies [4, 5, 6, 24, 53] focus on learning generic visual features by modeling the data distribution (e.”
Ieee Transactions on Pattern Analysis and Machine Intelligence 1 Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks (2014, unknown): “Training STL-10 CIFAR-10(400) Caltech-101 STL-10 67:1 AE 0:3 69:7 AE 0:3 7 9 :8 AE 0:5 CIFAR-10 64:5 AE 0:4 70:3 AE 0:4 77:8 AE 0:6 Caltech-101 66:2 AE 0:4 6 9 :5 AE 0:2 80:0 AE 0:5 Each column corresponds to different test data, each row to different training data (i.e., source of seed patches).”