Deep networks for robust visual recognition

By Yichuan Tang, Chris Eliasmith. 2010

104 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

Deep Learning: Methods and Applications (2014, methodology): “In [358], two strategies to improve the robustness of the DBN are developed.”
Deep learning for robust feature generation in audiovisual emotion recognition (2013, background): “Unsupervised Feature Learning and Deep Learning Deep learning techniques (See [9] for a survey) have become increasingly popular in various communities including speech and language processing [10] [11] [12] and vision processing [26] [27] [28] [29] [30].”
Event-driven contrastive divergence for spiking neuromorphic systems (2013, background): “One possible solution is to impose sparse connectivity between the layers [54, 41].”
An Overview of Deep-Structured Learning for Information Processing (2011, methodology): “In (Tang and Eliasmith, 2010), two strategies to improve the robustness of the DBN are developed.”
Unsupervised Spectral–Spatial Feature Learning With Stacked Sparse Autoencoder for Hyperspectral Imagery Classification (2015, background): “Such approaches learn feature hierarchies from unlabeled image data and have exhibited a very impressive performance on many recognition tasks such as multiple organ detection [12], handwritten digit recognition [13], and scene recognition [14].”
Deep learning for visual understanding: A review (2016, methodology): “To improve the robustness against occlusion and random noise, Lee et al. [63] applied two strategies: one is to take advantage of sparse connections in the ﬁ rst layer of the DBN to regularize the model, and the other is to develop a probabilistic de-noising algorithm.”
Generating more realistic images using gated MRF's (2010, methodology): “Following Gregor and LeCun [18] and also Tang and Eliasmith [19], we keep the number of parameters under control by using local potentials, but unlike Roth and Black [6] we only share weights between potentials that do not overlap.”
Dirty Pixels: Optimizing Image Classification Architectures for Raw Sensor Data (2017, background): “Several authors demonstrated that preprocessing noisy images with trained or classical denoisers improves the performance of trained classifiers [Tang and Eliasmith 2010; Tang et al. 2012; Agostinelli et al. 2013; Jalalvand et al. 2016; da Costa et al. 2016].”
Deep Learning and Its Applications to Signal and Information Processing (2011, methodology): “Tang and Eliasmith developed two strategies to improve the robustness of the DBN in [ 14 ].”
Deep learning for visual understanding (2017, unknown): “[89] applied two strategies: one is to take advantage of sparse connections in the rst layer of the DBN to regularize the model, and the other is to develop a probabilistic de-noising algorithm.”