Visual attention inspired distant view and close-up view classification

Song, Tong; Loh, Yuen Peng; Liang, Xuefeng; Kumada, Takatsune

doi:10.1109/icip.2016.7532867

Cited by 7 publications

(18 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(2) In order to classify the view-type, we model the focus and scale cues inspired by human visual attention, denoted as NSCT+SURF and AdobeBING+CNN respectively. This proposal shows superior performance over other baselines, with or without the HVS cues, and achieves the improvement from 84.00% in our prior framework [18] to 93.17%. (3) Due to the relatively new nature of the view-type classification problem in the field, we have also established a new benchmark containing 5050 natural narrow-view and wide-view images to facilitate our investigation and evaluate the proposed framework.…”

mentioning

confidence: 80%

“…In our previous work, we proposed a framework specifically designed with the inspiration from the visual attention in HVS [18], i.e. focus and scale cues, in performing view-type classification.…”

Section: Visual Attention Inspired Modelingmentioning

confidence: 99%

“…With the established frequency domain to emphasize the sharpness information of the images, the next step is to extract features to represent the different views. The Multi-Windows based Histogram of Frequency Energy (MWHFE) [18] approach quantifies the pixel-wise energies in the high-frequency components. Designed for the high-frequency domain, it captures the spatial distribution of the sharp details.…”

Section: ) Focus Feature Extraction and Classificationmentioning

confidence: 99%

“…We conducted exhaustive experiments to evaluate the performance of our visual attention inspired framework. The results show that the proposed framework greatly improves from the initial work [18]. Moreover, we also investigated algorithms from related domains that could be used for view-type classification, such as depth estimation [24], [25].…”

mentioning

confidence: 90%

“…Based on the two established cues, our initial work [18] modeled the focus cue and scale cue using Discrete Wavelet Transform (DWT) and Edge Boxes with Convolutional Neural Network (Edge Boxes+CNN), respectively. This prior work achieved a reasonable performance, however, our continuous investigation found that both the focus and scale cues are not optimally modeled with the current prototype.…”

mentioning

confidence: 99%

See 4 more Smart Citations

Wide or Narrow? A Visual Attention Inspired Model for View-Type Classification

et al. 2019

Self Cite

View full text Add to dashboard Cite

Emerging research revealed that the view-type of photos is not only related to the field of data sciences, such as the sentiment brought forth by sightseeing spots, but also in the field of social sciences about human emotions and behaviors. These potential usages of view-types trigger a challenging problem, that is to automatically distinguish them into wide or narrow. In this paper, we present a computational model to classify them inspired by the human visual system. We found two cues that can represent the visual attention, i.e., focus cue and scale cue. The focus cue is modeled in the frequency domain using the non-sampled contourlet transform (NSCT) and speeded up robust features (SURF). The scale cue is modeled by defining the spatial size and conceptual sizes of an object in the image, whereby AdobeBING and convolutional neural network are used for the respective measurements. By integrating these focus and scale models, a robust scheme is hence proposed for this non-trivial task. The experiments on a newly established dataset, which has 5050 natural images, show better performance by our proposal when compared to the state-of-the-arts.

show abstract

mentioning

confidence: 80%

Section: Visual Attention Inspired Modelingmentioning

confidence: 99%

Section: ) Focus Feature Extraction and Classificationmentioning

confidence: 99%

mentioning

confidence: 90%

mentioning

confidence: 99%

See 3 more Smart Citations

Wide or Narrow? A Visual Attention Inspired Model for View-Type Classification

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

Large Field/Close-Up Image Classification: From Simple to Very Complex Features

Le¹,

Ladret²,

Nguyen³

et al. 2019

Computer Analysis of Images and Patterns

View full text Add to dashboard Cite

In this paper, the main contribution is to explore three different types of features including Exchangeable Image File (EXIF) features, handcrafted features and learned features in order to address the problem of large field / close up images classification with a Support Vector Machine (SVM) classifier. The impacts of every feature set on classification performances and computational complexities are investigated and compared to each other. Results prove that learned features are of course very efficient but with a computational cost that might be unreasonable. On the contrary, it appears that it is worthy to consider EXIF features when available because they represent a very good compromise between accuracy and computational cost. Keywords: image classification • large field image • close-up image • handcrafted features • exif features • learned features • feature evaluation • feature selection • support vector machine • transfer learning.

show abstract