Spatio-temporal saliency based on rare model

Decombas, Marc; Riche, Nicolas; Dufaux, Frédéric; Pesquet‐Popescu, Béatrice; Mancaş, Matei; Gosselin, Bernard; Dutoit, Thierry

doi:10.1109/icip.2013.6738712

Cited by 11 publications

(4 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Characteristics: global | HL | both | color Authors: M. Decombas, N. Riche, F. Dufaux, B. Pesquet-Popescu, M. Mancas, B. Gosselin, and T. Dutoit [17]. Description: STRAP is a new saliency model based on a spatiotemporal rarity mechanism and integrating prior information.…”

Section: Strap: Spatiotemporal Multiscale Rarity Algorithm With Priormentioning

confidence: 98%

Bottom-Up Saliency Models for Videos: A Practical Review

Riche

Mancaş

2016

From Human Attention to Computational Attention

Self Cite

View full text Add to dashboard Cite

Research on visual saliency initially focused on still images rather than on video content. However, in the recent years, an increasing demand of video saliency appeared for some applications like gaming, editing, video retargeting, smart TV, robot navigation, surveillance, etc. Therefore, remarkable progress has been made first in the understanding on eye tracking data with dynamical stimuli and, in a second time, in the modeling process.There are fundamental differences between videos and still images. For example, each video frame is only observed during a fraction of a second, while a still image can be viewed much longer. Some videos can feature varying camera motion such as tilting, panning, zooming, etc. For this reason, videos are probably viewed differently by human observers than still images, and some comprehensive comparative studies have emerged. In [1], for example, the authors study the influence of tasks on gaze behavior in static and dynamic scenes. In [2], the gaze on static and dynamic scene is compared; it also shows that the center bias decreases with dynamic stimuli.In terms of modeling, static models have first been extended to video. This is the case for GBVS, SDSR, NMPT, or SSOI where authors added dynamic features to their models. Though these existing models are major contributions, video saliency estimation methods should then differ substantially from image saliency methods.

show abstract

Section: Strap: Spatiotemporal Multiscale Rarity Algorithm With Priormentioning

confidence: 98%

Bottom-Up Saliency Models for Videos: A Practical Review

Riche

Mancaş

2016

From Human Attention to Computational Attention

Self Cite

View full text Add to dashboard Cite

show abstract

“…1 presents the proposed approach of video summarization using seam carving with spatio-temporal grouping constraint. From an original video of length T, saliency maps are created with the ST -RARE model [17] computed in the (x,y) plane. This model identifies the salient objects by finding the rarity on different maps.…”

Section: A Review Of Seam Carvingmentioning

confidence: 99%

Spatio-temporal grouping with constraint for seam carving in video summary application

Decombas

Dufaux

Pesquet‐Popescu

2013

2013 18th International Conference on Digital Signal Processing (DSP)

Self Cite

View full text Add to dashboard Cite

Efficient tools to perform video summarization become essential as the number of cameras for video surveillance is growing exponentially. The development of automatic algorithms to aid human operators in identifying what is important and keeping it in a video summary becomes essential. This paper proposes a new way to obtain video summary based on seamcarving. An efficient spatio-temporal grouping is done to determine the temporal rate of reduction depending on the content, to suppress groups of isolated seams, to identify spatio

show abstract

“…Regions with high local contrast, global rare spatial or temporal features can be used for saliency detection as well [2]. The high contrast is not restricted to the spatial domain, it can also be extended to the temporal domain [3].…”

Section: Introductionmentioning

confidence: 99%

“…The rest of the variables and the function are defined in Equation (2). Lastly, the texture's entropy for feature index 14 is defined as:…”

mentioning

confidence: 99%