Deep networks for saliency detection via local estimation and global search

Abstract

This paper presents a saliency detection algorithm by integrating both local estimation and global search. In the local estimation stage, we detect local saliency by using a deep neural network (DNN-L) which learns local patch features to determine the saliency value of each pixel. The estimated local saliency maps are further refined by exploring the high level object concepts. In the global search stage, the local saliency map together with global contrast and geometric information are used as global features to describe a set of object candidate regions. Another deep neural network (DNN-G) is trained to predict the saliency score of each object region based on the global features. The final saliency map is generated by a weighted sum of salient object regions. Our method presents two interesting insights. First, local features learned by a supervised scheme can effectively capture local contrast, texture and shape information for saliency detection. Second, the complex relationship between different global saliency cues can be captured by deep networks and exploited principally rather than heuristically. Quantitative and qualitative experiments on several benchmark data sets demonstrate that our algorithm performs favorably against the state-of-theart methods.

Where Should Saliency Models Look Next?

“…Pan et al [22] also trained two architectures on SALICON in an end-to-end manner: a shallow convnet trained from scratch, and a deeper one whose first three layers were adapted from the VGG network (SalNet). Other saliency models based on deep learning have been proposed for salient region detection [23,24,25,26]. In this paper, we focus on predicting eye fixations rather than detecting and segmenting salient objects in scenes.…”

Saliency Aggregation: Multifeature and Neighbor Based Salient Region Detection for Social Images

“…However, TBS method uses object information. The same or similar [32], CA [33], CB [34], DRFI [25], DSR [35], FT [23], GC [36], GS [37], HM [38], HS [7], LEGS [10], LRR [39], MC [40], MCDL [9], MR [24], PCA [41], BD [42], RC [14], RFCN [11], SBF [43], SEG [44], SF [45], SMD [46], MDF [8], SS [47], SVO [17], TD [48], and DBS. objects can ensure similar salient regions to some extent.…”

“…OBS method was compared with 11 state-of-the-art methods, including FT [23], RC [14], SF [45], HS [7], MR [24], DRFI [25], GC [36], MC [40], BD [42], MDF [8], and LEGS [10].…”

“…DBS method was compared with 27 state-of-the-art methods in Section 7.2.1. 27 state-of-the-art methods are CB [34], FT [23], SEG [44], RC [14], SVO [17], LRR [39], SF [45], GS [37], CA [33], SS [47], HS [7], TD [48], MR [24], DRFI [25], PCA [41], HM [38], GC [36], MC [40], DSR [35], SBF [43], BD [42], SMD [46], BL [32], MCDL [9], MDF [8], LEGS [10], and RFCN [11]. These methods not only are very popular but also cover many types.…”

“…It has been found that only relying on low-level features cannot achieve satisfactory results. The researches have proved that the hierarchical and deep architectures [7][8][9][10][11][12] for salient region detection are very effective. Thus, a salient region detection method based on deep learning is proposed in this paper.…”