Action and Event Recognition with Fisher Vectors on a Compact Feature Set

By Dan Oneaţă, Jakob Verbeek, Cordelia Schmid. 2013

434 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

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

Delving into Egocentric Actions (2015, background): “IFV [26] has been shown to outperform other encoding methods in action recognition [24].”
P-CNN: Pose-Based CNN Features for Action Recognition (2015, methodology): “As in [25] we use a spatial pyramid representation and a number of K = 256 Gaussian components.”
Modeling Spatial-Temporal Clues in a Hybrid Deep Learning Framework for Video Classification (2015, methodology): “In addition, further improvements may be achieved by using advantageous feature encoding methods like the Fisher Vectors [29] or adopting feature normalization strategies, such as RootSift [1] and Power Norm [32].”
Completeness Modeling and Context Separation for Weakly Supervised Temporal Action Localization (2019, methodology): “Early methods were mainly based on hand-crafted features [26, 43, 34].”
Gaussian Temporal Awareness Networks for Action Localization (2019, background): “The main stream of content-independent proposal algorithms is uniformly or sliding window-ly sampling in a video [24, 33, 41], which leads to huge computations for further classification.”
DISCOVER: Discovering Important Segments for Classification of Video Events and Recounting (2014, background): “Recently, [15] showed that using multiple types of low level features can increase event classification performance.”
Feature sampling and partitioning for visual vocabulary generation on large action classification datasets (2014, background): “So far the best global representation for challenging video data has been in the form of a bag-of-features (BoF) histogram, and its extensions to Fisher vectors [2].”
Bags of Spacetime Energies for Dynamic Scene Recognition (2014, background): “Recently, the Fisher vector representation [19] has shown state-of-the-art results for a variety of visual tasks (e.g., [5, 14, 18, 23]); in contrast, here it is found that locality constrained linear coding [27] performs particularly well for dynamic scenes.”
Spatio-temporal Object Detection Proposals (2014, positive): “The second important factor is the use of state-of-the-art local feature pooling methods, such as the Fisher vector [28,36].”
Spatial-Aware Object Embeddings for Zero-Shot Localization and Classification of Actions (2017, positive): “Such representations include local spatio-temporal interest points and features [26, 31, 52] and local trajectories [1, 50], typically aggregated into VLAD or Fisher vector representations [38, 39].”
A discriminative CNN video representation for event detection (2014, background): “Recent research efforts have shown that combining multiple features, including static appearance features [9, 25, 41], motion features [23, 7, 43, 44, 33] and acoustic features [28], yields good performance in event detection, as evidenced by the reports of the top ranked teams in the TRECVID Multim”
The THUMOS challenge on action recognition for videos "in the wild" (2016, positive): “…in large part by the release of large video datasets such as UCF101 ( Soomro et al., 2012 ) and HMDB ( Kuehne et al., 2011 ) in conjunction with the development of new features ( Wang and Schmid, 2013 ), representations ( Oneata et al., 2013 ) and learning methods ( Simonyan and Zisserman, 2014a ).”
LEAR-INRIA submission for the THUMOS workshop (2013, background): “In recent evaluations [2, 7], this shows an improved performance over bag of features for both image and action classification.”
Temporal Action Detection with Structured Segment Networks (2017, methodology): “Previous works on activity detection mainly use sliding windows as candidates and focus on designing hand-crafted feature representations for classiﬁcation [11, 44, 30, 25, 57, 18].”
The LEAR submission at Thumos 2014 (2014, methodology): “Following [7], we re-score the detection windows by multiplying the detection score by the duration of the window.”