Real-time video summarization on mobile

Marvaniya, Smit; Damoder, Mogilipaka; Gopalakrishnan, Viswanath; Iyer, Kiran Nanjunda; Soni, Kapil Dev

doi:10.1109/icip.2016.7532342

Cited by 8 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Rushes videos are summarized in an online fashion in [170], by constructing a decision tree and evaluating the cumulative frame importance over a fixed time period. On the other hand, generic proposals in this category involve creating an online dictionary [106,201] to discriminate among redundant and unique segments. Optimization based techniques are suggested in [35] for identifying the representative segments in an iterative way.…”

Section: Onlinementioning

confidence: 99%

See 1 more Smart Citation

Video Skimming: Taxonomy and Comprehensive Survey

K.,

Sen,

Raman

2019

Preprint

View full text Add to dashboard Cite

Video skimming, also known as dynamic video summarization, generates a temporally abridged version of a given video. Skimming can be achieved by identifying significant components either in uni-modal or multimodal features extracted from the video. Being dynamic in nature, video skimming, through temporal connectivity, allows better understanding of the video from its summary. Having this obvious advantage, recently, video skimming has drawn the focus of many researchers benefiting from the easy availability of the required computing resources. In this paper, we provide a comprehensive survey on video skimming focusing on the substantial amount of literature from the past decade. We present a taxonomy of video skimming approaches, and discuss their evolution highlighting key advances. We also provide a study on the components required for the evaluation of a video skimming performance.

show abstract

Section: Onlinementioning

confidence: 99%

“…Selection of group of frames for a summary is done based on setting a threshold for the distance. A real time video summarization on mobile devices is suggested in [106]. Video is segmented with the help of motion as in [53].…”

Section: Onlinementioning

confidence: 99%

Video Skimming: Taxonomy and Comprehensive Survey

K.,

Sen,

Raman

2019

Preprint

View full text Add to dashboard Cite

show abstract

“…Video summarization. The goal of summarization techniques [10,11,12] is to generate a shorter version of the video keeping the essential information by either creating a static storyboard or still-image abstract, where some selected frames resume the relevant video content [13], or a dynamic video skimming or moving-image abstract, where selected clips from the original stream are collated to compose the output video [14,15]. Molino et al .…”

Section: Related Workmentioning

confidence: 99%

Making a long story short: A multi-importance fast-forwarding egocentric videos with the emphasis on relevant objects

Silva¹,

Ramos²,

Cadar³

et al. 2018

Journal of Visual Communication and Image Representation

View full text Add to dashboard Cite

The emergence of low-cost high-quality personal wearable cameras combined with the increasing storage capacity of video-sharing websites have evoked a growing interest in first-person videos, since most videos are composed of longrunning unedited streams which are usually tedious and unpleasant to watch. State-of-the-art semantic fast-forward methods currently face the challenge of providing an adequate balance between smoothness in visual flow and the emphasis on the relevant parts. In this work, we present the Multi-Importance Fast-Forward (MIFF), a fully automatic methodology to fast-forward egocentric videos facing these challenges. The dilemma of defining what is the semantic information of a video is addressed by a learning process based on the preferences of the user. Results show that the proposed method keeps over 3 times more semantic content than the state-of-the-art fast-forward. Finally, we discuss the need of a particular video stabilization technique for fast-forward egocentric videos 1 .

show abstract

“…Examples of these methodologies are the use of GIST difference over a given window [7]; similarity between the R-CNN hashes extracted from the fixation region [132]; the Cumulative Displacement Curves in [100]; a Super Vector Machine-Hidden Markov Model pipeline in [126]; a 3D Deep Convolutional Neural Network in [102]; and KTS by Potapov et al [103], a kernelbased change point detection algorithm. Recently, many authors chose to segment the video deterministically, set to a specific number of frames or time [48,54,80,86,93,99,108,109,130,131,145].…”

Section: Event Segmentationmentioning

confidence: 99%

“…Alternatively, the presence of factors such as saliency, edges and colorfulness, object interactions, and the presence of landmarks, people or faces can be used to predict each frames' interestingness [47]. Similarly, [7,62,86,129] base their quality or intentionality prediction on cues for composition such as picture alignment or accelerometer data; the artistic rule-of-thirds; symmetry (on local SIFT features) and color vibrancy; head tilt; camera motion; etc. Context dependent, Okamoto et al [93] train a model for navigation instructions using a crosswalk detector and ego-motion cues.…”

Section: Supervised Learningmentioning

confidence: 99%

Visual understanding and personalization for an optimal recollection experience

Ana¹

View full text Add to dashboard Cite

show abstract

Real-time video summarization on mobile

Cited by 8 publications

References 17 publications

Video Skimming: Taxonomy and Comprehensive Survey

Video Skimming: Taxonomy and Comprehensive Survey

Making a long story short: A multi-importance fast-forwarding egocentric videos with the emphasis on relevant objects

Visual understanding and personalization for an optimal recollection experience

Contact Info

Product

Resources

About