Perception-oriented video saliency detection via spatio-temporal attention analysis

Zhong, Sheng-hua; Liu, Yan; Ng, To-Yee; Liu, Yang

doi:10.1016/j.neucom.2016.04.048

Cited by 14 publications

(5 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Reference [22] puts forward that in the training process of sports video panorama synthesis technology, panorama synthesis technology can effectively play its role as a media platform and help students understand the principle of sports video panorama synthesis technology more deeply in various forms such as sound, image, graphics, video, or animation. Reference [23] through the method of big data analysis, in the process of sports training, coaches and athletes do static analysis on the completion of movements, master the essentials of technical movements from the overall situation, realize the essentials of coherent movements, correct wrong movements and wrong timing of exertion, reduce blind repetition, improve training efficiency, and reduce the possibility of injury to athletes. Reference [24] research shows that in the process of sports video, the athletes' skills and movements in different stages and scenes are simply superimposed on the static scenes through the video panorama, and the athletes' movement tracks and technical details in the practice process can be fully displayed with the help of frozen work effects.…”

Section: Related Workmentioning

confidence: 99%

Sports Video Panorama Synthesis Technology Based on Edge Computing and Video Shot Boundary Detection

Chen

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

The development requirements for sports panorama synthesis technology, which rely on modern network technology, abandon traditional basketball training forms, and make effective use of the application and development of video panorama technology in the actual training process, are not only an important response to the current characteristics of students’ physical education learning and physical and mental growth but also an important response to the current characteristics of students’ physical education learning and physical and mental growth. In sports video analysis technology, a sports video panorama is a technical tool that converts an action video into a static action panorama to achieve the effect of action freezing and make overall analysis and mastery of the action easier. Shot segmentation is the foundation of hierarchical video structure, and it necessitates the accurate detection of all types of complex edited shot boundaries, as well as the effective distinction of motion changes in shots, in order to avoid shot boundary recognition being hampered. The synthesis technology for sports video panorama is investigated in this paper using edge computing and video shot boundary detection. After obtaining the boundary feature that describes the video shot, a comparison of this feature with a predetermined threshold value can be used to determine whether there is shot shear.

show abstract

Section: Related Workmentioning

confidence: 99%

Sports Video Panorama Synthesis Technology Based on Edge Computing and Video Shot Boundary Detection

Chen

2022

Wireless Communications and Mobile Computing

View full text Add to dashboard Cite

show abstract

“…Attention mechanism can be described as a weighted sum of values, where the weights assigned to each value are computed by a compatibility function [5]. Attention mechanism has become more and more popular in various tasks such as recommender system [6],machine translation [11] and Multimedia [12]- [14]. Recently, Transformer [5] was proposed and achieved promising empirical results in machine translation.…”

Section: A Attention Mechanismmentioning

confidence: 99%

Deep Self-Attention for Sequential Recommendation (S)

Zhang¹

2021

Proceedings of the 33rd International Conference on Software Engineering and Knowledge Engineering

View full text Add to dashboard Cite

Sequential recommendation aims to recommend the next item that a user will likely interact with by capturing the useful sequential patterns from users' historical behaviors. Recently, it has become an important and popular component in various e-commerce platforms. As a successful network, Transformer has been widely used to adaptively capture the dynamics of users' historical behaviors for sequential recommendation. In recommender systems, the size of embedding is usually set to be small. Under small embedding, the dot-product in Transformer may have the limitation on calculating the complex relevance between keys and queries. To address the common but neglected issue, in this paper, we present a new model, Deep Self-Attention for Sequential Recommendation (DSASrec), which proposes a chunking deep attention to compute attention weights. The chunking deep attention has two modules: a deep module and a chunking module. The deep module is used to improve the nonlinearity of the attention function. The chunking module is used to calculate attention weights several times like the multihead attention in Transformer. Extensive experiments on three benchmark datasets show that our model can achieve state-ofthe-art results. Our implementation is available in PyTorch 1 .

show abstract

“…While bottom-up attention modelling, i.e. solely using image features, has been extensively studied in controlled laboratory settings, egocentric settings are characterised by a mix [58]. Saliency models, which aim to predict which image regions most attract viewers' attention are an important type of computational model of visual attention [18].…”

Section: Computational Modelling Of Egocentric Attentionmentioning

confidence: 99%

Forecasting user attention during everyday mobile interactions using device-integrated and wearable sensors

Steil

Müller

Sugano

et al. 2018

Proceedings of the 20th International Conference on Human-Computer Interaction With Mobile Devices and Services

View full text Add to dashboard Cite

Visual attention is highly fragmented during mobile interactions, but the erratic nature of attention shifts currently limits attentive user interfaces to adapting after the fact, i.e. after shifts have already happened. We instead study attention forecasting -the challenging task of predicting users' gaze behaviour (overt visual attention) in the near future. We present a novel long-term dataset of everyday mobile phone interactions, continuously recorded from 20 participants engaged in common activities on a university campus over 4.5 hours each (more than 90 hours in total). We propose a proof-of-concept method that uses device-integrated sensors and body-worn cameras to encode rich information on device usage and users' visual scene. We demonstrate that our method can forecast bidirectional attention shifts and predict whether the primary attentional focus is on the handheld mobile device. We study the impact of different feature sets on performance and discuss the significant potential but also remaining challenges of forecasting user attention during mobile interactions.

show abstract

Perception-oriented video saliency detection via spatio-temporal attention analysis

Cited by 14 publications

References 54 publications

Sports Video Panorama Synthesis Technology Based on Edge Computing and Video Shot Boundary Detection

Sports Video Panorama Synthesis Technology Based on Edge Computing and Video Shot Boundary Detection

Deep Self-Attention for Sequential Recommendation (S)

Forecasting user attention during everyday mobile interactions using device-integrated and wearable sensors

Contact Info

Product

Resources

About