Spatio-temporal combination of saliency maps and eye-tracking assessment of different strategies

Chamaret, Christel; Chevet, Jean-Claude; Meur, Olivier Le

doi:10.1109/icip.2010.5651381

Cited by 29 publications

(12 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Due to these limitations, these models do not perform well on new and streaming real-world data. Some of these models have been compared against human eye tracking data [3,7,[9][10]. As expected, there is low correlation between human fixation and models.…”

Section: Introductionmentioning

confidence: 92%

Brain-enhanced synergistic attention (BESA)

Khosla

Keegan

Zhang

et al. 2012

Proceedings of the 4th Workshop on Eye Gaze in Intelligent Human Machine Interaction

View full text Add to dashboard Cite

In this paper 1 , we describe a hybrid human-machine system for searching and detecting Objects of Interest (OI) in imagery. Automated methods for OI detection based on models of human visual attention have received much interest, but are inherently bottom-up and driven by features. Humans fixate on regions of imagery based on a much stronger top-down component. While it may be possible to include some aspects of top-down cognition into these methods, it is difficult to fully capture all aspects of human cognition into an automated algorithm. Our hypothesis is that combination of automated methods with human fixations will provide a better solution than either alone. In this work, we describe a Brain-Enhanced Synergistic Attention (BESA) system that combines models of visual attention with real-time eye fixations from a human for accurate search and detections of OI. We describe two different BESA schemes and provide implementation details. Preliminary studies were conducted to determine the efficacy of the system and initial results are promising. Typical applications of this technology are in surveillance, reconnaissance and intelligence analysis.

show abstract

Section: Introductionmentioning

confidence: 92%

Brain-enhanced synergistic attention (BESA)

Khosla

Keegan

Zhang

et al. 2012

Proceedings of the 4th Workshop on Eye Gaze in Intelligent Human Machine Interaction

View full text Add to dashboard Cite

show abstract

“…Different fusion strategies can be utilized [12], such as additive fusion, multiplicative fusion and maximum fusion. In our DCVS codec, we simply adopt the maximum fusion to obtain the spatiotemporal saliency (i.e., S(f SI ) = max(S s (f SI ), S t (f SI ))) because this fusion method can adequately detect most of the salient regions in the SI frame and circumvent the difficulty of determining the fusing weights for spatial and temporal saliency.…”

Section: B Perceptually-aware Bcs For a Non-key Framementioning

confidence: 99%

Perceptually-aware distributed compressive video sensing

Djahel

Qiao

et al. 2015

2015 Visual Communications and Image Processing (VCIP)

View full text Add to dashboard Cite

By combining the advantages of distributed video coding (DVC) and compressive sensing (CS), distributed compressive video sensing (DCVS) poses itself as a very promising lowcomplexity video coding framework for distributed applications. In order to improve the rate-distortion performance of DCVS, much research efforts have been focused on exploring the best ways to utilize the spatial/temporal redundancy of video data to achieve efficient sparse representation and reconstruction at the decoder. Unlike the existing DCVS schemes, we aim to improve the perceptual rate-distortion performance of DCVS by designing a novel perceptually-aware DCVS codec. Based on online estimation of the correlation noise between a nonkey frame and its side information (SI) considering the effect of human visual system (HVS), we design an efficient perceptuallyaware block compressive sensing scheme for a non-key frame in our DCVS codec, in order to more accurately reconstruct the salient regions in the video frames. The obtained experimental results reveal that our DCVS codec outperforms the legacy DCVS codecs in terms of the perceptual rate-distortion performance.

show abstract

“…Dealing with different sources of information, their correct combination is a great challenge [7]. In this work we propose to fuse several conspicuity maps considering the human visual characteristics.…”

Section: Neural Network For the Combination Of Featuresmentioning

confidence: 99%

Visual attention modeling for 3D video using neural networks

Iatsun¹,

Larabi²,

Fernández-Maloigne³

2014

2014 International Conference on 3D Imaging (IC3D)

View full text Add to dashboard Cite

International audienceVisual attention is one of the most important mechanisms in the human visual perception. Recently, its modeling becomes a principal requirement for the optimization of the image processing systems. Numerous algorithms have already been designed for 2D saliency prediction. However, only few works can be found for 3D content. In this study, we propose a saliency model for stereoscopic 3D video. This algorithm extracts information from three dimensions of content, i.e. spatial, temporal and depth. This model benefits from the properties of interest points to be close to human fixations in order to build spatial salient features. Besides, as the perception of depth relies strongly on monocular cues, our model extracts the depth salient features using the pictorial depth sources. Since weights for fusion strategy are often selected in ad-hoc manner, in this work, we suggest to use a machine learning approach. The used artificial Neural Network allows to define adaptive weights based on the eye-tracking data. The results of the proposed algorithm are tested versus ground-truth information using the state-of-the-art techniques

show abstract

Spatio-temporal combination of saliency maps and eye-tracking assessment of different strategies

Cited by 29 publications

References 6 publications

Brain-enhanced synergistic attention (BESA)

Brain-enhanced synergistic attention (BESA)

Perceptually-aware distributed compressive video sensing

Visual attention modeling for 3D video using neural networks

Contact Info

Product

Resources

About