Large-Scale Optimization of Hierarchical Features for Saliency Prediction in Natural Images

By E. Vig, Michael Dorr, David D. Cox. 2014

451 citations recorded by the source corpus.

Subfields: Cognitive Neuroscience, Computer Vision and Pattern Recognition, Sensory Systems

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Computer Vision and Pattern Recognition

In Sensory Systems

Citations in the corpus

Recent advances in convolutional neural networks (2015, methodology): “In [171], an ensemble of CNNs is derived from a large number of randomly instantiated CNN models, to generate good features for saliency detection.”
Faster gaze prediction with dense networks and Fisher pruning (2018, methodology): “Vig et al. [30] trained a smaller network end-to-end but did not start from a pretrained representation and therefore did not achieve the performance of current state-of-the-art architectures.”
TurkerGaze: Crowdsourcing Saliency with Webcam based Eye Tracking (2015, methodology): “We compared the saliency map based on AMTurk subjects’ fixations to various state-of-the-art computer models of saliency[36, 43, 15] and the Judd inter-subject-agreement baseline.”
Deep Gaze I: Boosting Saliency Prediction with Feature Maps Trained on ImageNet (2014, methodology): “The three top performing models after Deep Gaze I are in order of decreasing performance: BMS (82.57%, Zhang and Sclaroﬀ, 2013), Mixture of Saliency Models (82.09%, Han and Satoh, 2014), and eDN (81.92%, Vig et al., 2014).”
SalGAN: Visual Saliency Prediction with Generative Adversarial Networks (2017, background): “The Ensemble of Deep Networks (eDN) [30] represented an early architecture that automatically learns the representations for saliency prediction, blending feature maps from different layers.”
Paying More Attention to Saliency: Image Captioning with Saliency and Context Attention (2017, background): “Firstattempts of computing saliency maps through a neural network lacked from the absence of sufficiently large training datasets [27, 32, 48].”
Recurrent Attentional Networks for Saliency Detection (2016, unknown): “Recently, deep learning-based saliency detection methods [44, 47, 49, 29, 43] have been very successful.”
A deep multi-level network for saliency prediction (2016, background): “A ﬁrst attempt to model saliency with deep convolutional networks (DCNs) has been recently proposed by Vig et al. [10].”
Revisiting Video Saliency: A Large-Scale Benchmark and a New Model (2018, background): “Deep learning based static saliency models [54, 35, 24, 39, 47, 29, 56, 57] have achieved astonishing improvements, relying on the powerful end-to-end learning ability of neural network and the availability of large-scale static saliency datasets [31].”
Predicting Human Eye Fixations via an LSTM-Based Saliency Attentive Model (2016, methodology): “Unlike previous approaches that include handcrafted priors [10], [23], [12], [24], [25], our module keeps the architecture trainable end-toend and can learn priors in an automatic way.”
A Deep Network Solution for Attention and Aesthetics Aware Photo Cropping (2019, methodology): “13 ATC [16] + eDN [30] (MaxAvg) 0.”
Spatiotemporal Knowledge Distillation for Efficient Estimation of Aerial Video Saliency (2019, methodology): “shallow-learning saliency models [29], [30], [31], [32] aim to directly learn an optimal fusion strategy of hand-crafted features from data.”
RGBD Salient Object Detection via Deep Fusion (2016, methodology): “As in the proposed method, CNNs have been adopted in several other works to extract hierarchical feature representations for detecting salient regions [38], [39], [40], [41], [42], [43].”
Object Detection With Deep Learning: A Review (2018, methodology): “[28] follows a completely automatic data-driven approach to perform a largescale search for optimal features, namely an ensemble of deep”
End-to-end Convolutional Network for Saliency Prediction (2015, methodology): “An early attempt of predicting saliency model with a convnet was the ensembles of Deep Networks (eDN) [19], which proposed an optimal blend of features from three different convnet layers who were finally combined with a simple linear classifier trained with positive (salient) or negative (non-salie”
SALICON: Reducing the Semantic Gap in Saliency Prediction by Adapting Deep Neural Networks (2015, methodology): “Typically, a Support Vector Machine (SVM) is used [21, 33, 39, 40].”
Deep Visual Attention Prediction (2017, methodology): “…Yes Yes Mr-CNN [1] 400 × 400 Yes Yes SalNet [15] 320 × 240 Yes Yes Deep Gaze I [42] full size Yes Yes Saliency Uniﬁed [16] 417 × 417 Yes Yes eDN [13] 512 × 384 Yes Yes AIM [56] full size Yes No Judd Model [24] 200 × 200 Yes No BMS [57] w = 600 No No CAS [41] max { w,h } =250 No No GBVS [38]…”