Video summarization by learning submodular mixtures of objectives

By Michael Gygli, Helmut Grabner, Luc Van Gool. Computer Vision and Pattern Recognition, 2015

428 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/cfcb9bcc1e8b4d3451578398aeb37f0fa5614632

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Subset Replay Based Continual Learning for Scalable Improvement of Autonomous Systems (2018, methodology): “This has been successfully used in video summarization [11] where the focus is to come up with a succinct yet informative summary of otherwise long videos.”
Summarization of Egocentric Videos: A Comprehensive Survey (2017, methodology): “Some works already use NLP measures to evaluate their summaries, as is the case of [22], to support their f -measure evaluation, and [21].”
Deep Attentive Video Summarization With Distribution Consistency Learning (2020, methodology): “[28] formulated video summarization as a subset selection problem by learning submodular mixtures of objectives for different criteria directly.”
Video Summarization Using Deep Neural Networks: A Survey (2021, methodology): “splits; 5-fold cross validation), the way that the computed FScores from the pairwise comparisons with the different user summaries are taken under consideration (maximum value is kept for SumMe according to [100]; average value is kept for TVSum) to form the F-Score for a given test video, and the ”
Textually Customized Video Summaries (2017, methodology): “Supervised methods use summary annotations obtained from human workers in the form of ground truth summaries [7, 8], highlight annotations [45] or GIF-formatted summaries [9].”
Enhancing Video Summarization via Vision-Language Embedding (2017, methodology): “We use the same method as in [9] to produce a per-frame interestingness score for all the frames in a video segment.”
A Weighted Sparse Sampling and Smoothing Frame Transition Approach for Semantic Fast-Forward First-Person Videos (2018, background): “Techniques typically use features that range from low-level such as motion and color [32, 5] to high-level (e.”
Unsupervised Video Summarization with Adversarial LSTM Networks (2017, methodology): “For datasets with multiple human annotations (in the form of key shots or key frames), we follow the standard approach described in [11, 34, 43] to create a single groundtruth set for evaluation.”
Video Summarization With Attention-Based Encoder–Decoder Networks (2017, methodology): “For instance, Gygli et al. [15] learn to combine the criteria of representative, relevance, and uniformity to ensure the generated summaries are the most consistent with the reference ones.”
SCSampler: Sampling Salient Clips From Video for Efficient Action Recognition (2019, background): “For example, video summarization [18, 19, 29, 37, 56, 57, 58] and the automatic production of sport highlights [30, 31] entail creating a much shorter version of the original video by concatenating a small set of snippets corresponding to the most informative or exciting moments.”
A Practical Online Framework for Extracting Running Video Summaries under a Fixed Memory Budget (2021, methodology): “For example, Gygli et al. [12] learn a mixture of components to model the objective for video summarization in the oﬄine setting.”
Making 360° Video Watchable in 2D: Learning Videography for Click Free Viewing (2017, background): “Video summarization Video summarization aims to generate a concise representation for a video by removing temporal redundancy while preserving the important events [8, 14–16, 24, 27, 32, 33, 38, 42].”
Dense-Captioning Events in Videos (2017, methodology): “In an effort to start describing videos, methods in video summarization aimed to congregate segments of videos that include important or interesting visual information [62, 59, 16, 3].”
Localizing Moments in Video with Natural Language (2017, methodology): “Though most summarization algorithms do not include textual input ([3, 9, 10, 52, 53]), some use text in the form of video titles [21, 41] or user queries in the form of category labels to guide content selection [37].”
Viewpoint-Aware Video Summarization (2018, background): “However, owing to the subjective property of this task, a supervised approach [21, 38, 32, 23, 12, 31, 13, 19, 9, 42], that trains a summarization model which takes human-created summaries as the supervision, became standard because of its better performance.”
Video Summarization with Long Short-Term Memory (2016, methodology): “For example, a prior supervised approach learns to combine multiple hand-crafted criteria so that the summaries are consistent with ground truth [15,17].”
CLIP-It! Language-Guided Video Summarization (2021, methodology): “This enabled automatic evaluation of video summarization techniques and has lead to the development of many supervised learning based methods [6, 8, 23, 27, 32, 31, 41, 42, 43, 44, 45].”
Deep Reinforcement Learning for Unsupervised Video Summarization with Diversity-Representativeness Reward (2017, methodology): “Gygli et al. (Gygli et al. 2014) learned a linear regressor to predict the degree of interest-ingness of video frames and selected keyframes with the highest interestingness scores.”